We are making accurate observations of the changes in Doppler shift of stellar absorption lines. The scrambling of incident light by an optical fiber and the stability of wavelength calibration by a tilt-tunable Fabry-Perot etalon provide immunity to systematic errors. A cross-dispersed echelle spectrograph spatially separates the orders of constructive interference transmitted through the etalon. Selecting several echelle diffraction orders in the vicinity of 4250-4750 A, which are imaged on a CCD, about 350 Points on the profile of the stellar spectrum are sampled by successive orders of interferometric transmission through the etalon. At 4300 A each interference order is 47 milliangstroms wide and the sample points are 0.64 A apart, causing distinct, widely-spaced monochromatic images of the entrance aperture to be formed in the focal plane of the camera. Changes in Doppler shift modify the relative intensities of these images, in proportion to the slope of the spectral profile at each point sampled; for simplicity, the data are being analyzed only for changes in radial velocity rather than its absolute magnitude. With an argon emission line lamp the interferometer is calibrated to better than 2 parts in 100 million; this corresponds to + 6 meters/sec in Doppler shift. These calibrations show instrumental variations of + 12 meters/sec on a time scale of months; observations of stars are corrected for such changes. The internal repeatability of observations of the differential Doppler shift of Alpha Tau (K5 III; B=2.4) is + 6 meters/sec for each exposure of 224 square meter-seconds, and + 8 m/s for Beta Gem (KO III, B = 2.1). These exposures are obtained in 7 minutes with a 0.91-meter telescope. The external repeatability (day-to-day differential accuracy) of nightly averages of stellar observations is + 20 meters/second; improvements to the instrument are planned to reduce this value.

For high resolution spectroscopic observations in infrared astronomy, liquid helium cooled Fabry-Perot spectrometers have been developed. Making the spectrometers as simple and sturdy as possible, we have realized compact and stable spectrometers useful for even harsh conditions as in balloon or space observations as well as in ground based observations.

We have developed Fabry-Perot and Michelson based Interferometric Stellar Oscillation Spectrometers (FP-ISOS and Michelson-ISOS) to measure radial velocity curves of ultra-short period variable stars and to search for solar-type stellar oscillations. Both instruments are intrinsically stable at the 1 ms 1 level and with Michelson-ISOS we have achieved a photon shot noise limited sensitivity of better than 1 ms 1 on bright stars.

We have designed and fabricated a solid etalon from a 50-mm diameter wafer of optical-quality lithium niobate. The finished etalon has a free spectral range of 0.325 nm at 588 nm. The parallel faces are coated with silver, and the central 15-mm aperture of the etalon has a finesse of 18.6. The reflective faces double as electrodes, and application of voltage will shift the passband. This feature was used in a servo circuit to stabilize the passband' againsttemperature and tilt-induced drifts to better than three parts in 10. The etalon will be incorporated into a servo-controlled, tunable filter at the APL Solar Observatory. Because of its large aperture and high acceptance angle, it does not require that the host telescope or the filter be scanned mechanically in the course of normal solar measurements. Operated in the stabilized mode for day-long sessions, this filter alternately samples the wings of a narrow atomic absorption line in the solar spectrum and produces a signal proportional to velocity on the solar disk. The Fourier transform of this signal yields information on acoustic waves in the solar interior.

This paper presents a novel technique developed to increase the throughput of a Fabry-Perot spectrometer using a multiple zone aperture -(MZA). The method can very advantageously be used for study of line profiles from faint, extended objects. The description includes the use of a centrally masked MZA which, apart from increasing the throughput of the spectrometer, can suppress the radiation from any continuum source that may be present at the center. Detection of nebulosity around stars is one important area which can greatly be benefitted by this technique. The main hindrances in the detection of nebulosity are: (i) the faintness of the nebulosity, (ii) the uncertainty in its location and (iii) the stellar continuum radiation. Since the use of a centrally masked MZA (i) increases the throughput of the spectrometer, (ii) makes the search area circularly symmetric and (iii) suppresses the stellar continuum radiation, this forms an ideal technique for the purpose. The method has successfully been tested in the laboratory and the results of the laboratory tests are also presented.

The CCD/Transit Instrument (CTI) is a unique, fully-automated imaging survey telescope now in routine operation on Kitt Peak Mountain. This telescope, which has no moving parts, uses two charge-coupled devices (CCDs) aligned east-west in the focal plane. The CCDs are operated in the time-delay and integrate (TDI) mode at the apparent sidereal rate to produce a strip image of the sky 8.25 arcminutes wide in declination and eight hours of time in length. In this way, about 15 square degrees of the sky are surveyed every clear night. After a year, more than 40 square degrees will be surveyed in a continuous strip. Every clear night the CTI produces a V-bandpass image from one CCD to a limiting magnitude of V = 21 while the other CCD observes in the U, B, R or I bandpass, depending upon the sky brightness. Over several years, during which the telescope will not be moved, the digital CCD data can be registered and added to produce an image to fainter limiting magnitudes. Astronomical programs addressed by the CTI include the refinement of the supernova production rate by simply counting them as they occur over a number of years, the detection on the rise to maximum of several bright (Vmax < 16) supernovae per year, and the definition of a complete sample of quasars to faint limiting magnitude using as the primary selection criterion V-bandpass variability, with color as a secondary criterion. Other scientific programs addressed by CTI data include galactic structure investigations based on multi-color star counts to faint limiting magnitude, searching for solar/stellar variability at the millimagnitude level by precision photometry of the brightest unsaturated stars, galaxy counts and correlation functions in an "image without edges" (in one dimension), and the investigation of the statistics and galactic distribution of variable stars of all types. The CTI is now in routine operation. We report on the performance of its unique bi-metallic thermally self-compensating structure, its three-mirror wide-field optical system and its mode of operation. We also present an overview of the CTI data-handling system which is generating and absorbing up to 450 megabytes of image data on a daily basis and updating data bases at this rate. Both the real-time acquisition computer system and the reduction, analysis and archival computer systems are described. The role of the CTI as a prototype for future specialized, low-cost telescopes is discussed. Implications for successfully handling the vast amount of data produced by CTI for future large telescope data systems are reviewed.

A highly structured but extremely flexible database definition and interrogration system is described. This system is in daily use for reducing, archiving and evaluating data from the CCD/Transit Instrument (CTI) (McGraw et al. 1986, Cawson et al. 1986). Adopting a database system for managing observations and extracting scientific information from them has yielded many benefits of great use to astronomers: • a very few general purpose programs take the place of a multitude of special purpose programs, thus increasing reliability and reducing software effort and debugging time; • a powerful interrogation language allows interaction with the data in graphical forms and the definition of mathematical, relational and/or logical tests on the data during the interrogation as seems appropriate at the time; • the same command language can prompt for all parameters and supply 'reasonable' defaults making the database information easily accessible to collaborators who do not have to invest undue time 'learning the system'; • compatible databases can be produced, for instance to evaluate the structural properties of the telecope and the performance of the reduction software itself; and to test theoretical models using simulations, thereby unifying our engineering, software, theoretical and observational endeavors-in astronomy. Implications for the design of the future software systems for maintaining astronomical databases both at the national and at the individual astronomer levels are discussed. These impliations have the potential for radically changing the manner in which astronomers conceptualize their data.

The automated software system for archiving, analysing and interrogating data from the CCD/Transit Instrument (CTI) is described. The CTI collects up to 450 Mbytes of image-data each clear night in the form of a narrow strip of sky observed in two colors. The large data-volumes and the scientific aims of the project make it imperative that the data are analysed within the 24-hour period following the observations. To this end a fully automatic and self-evaluating software system has been developed. The data are collected from the telescope in real-time and then transported to Tucson for analysis. Verification is performed by visual inspection of random subsets of the data and obvious cosmic rays are detected and removed before permanent archival is made to optical disc. The analysis phase is performed by a pair of linked algorithms, one operating on the absolute pixel-values and the other on the spatial derivative of the data. In this way both isolated and merged images are reliably detected in a single pass. In order to isolate the latter algorithm from the effects of noise spikes a 3x3 Hanning filter is applied to the raw data before the analysis is run. The algorithms reduce the input pixel-data to a database of measured parameters for each image which has been found. A contrast filter is applied in order to assign a detection-probability to each image and then x-y calibration and intensity calibration are performed using known reference stars in the strip (e.g. AGK-3 stars). These are added to as necessary by secondary standards boot-strapped from the CTI data itself. The final stages involve merging the new data into the CTI Master-list and History-list and the automatic comparison of each new detection with a set of pre-defined templates in parameter-space to find interesting objects such as supernovae, quasars and variable stars. Each stage of the processing from verification to interesting image selection is performed under a data-logging system which both controls the pipe-lining of data through the system and records key performance monitor parameters which are built into the software. Furthermore, the data from each stage are stored in databases to facilitate evaluation, and all stages offer the facility to enter keyword-indexed free-format text into the data-logging system. In this way a large measure of certification is built into the system to provide the necessary confidence in the end results.

Current and future ground- and space-based telescopes are capable of producing immense volumes of data which, to make the telescope truly productive, must be analyzed in reasonable periods of time after acquisition. Classical methods of computerized data analysis tend to become more specific as the data volume increases, thereby not fully utilizing the information content of the data. To reverse this trend, we investigate artificial intelligence-based analysis systems to act either as the principal analysis tool for a data set, or to act as a co-processor to a classical statistical analysis system. We have designed, and are in the process of implementing, an image processing system based on concepts of artificial intelligence. The input images are produced by the CCD Transit Instrument (CTI)1. Standard astronomical classification has generally been accomplished by comparison to a hierarchy of standard objects (e.g. the MK spectral classification system2) and, in our new system, we mimic the use of such standards by a network of prototypes. The prototypes are represented within the computer as frames3, each of which contains knowledge either of a standard object or of the links between such objects. Thus the frames provide both the goal of our classification search, and the information required to guide that search. Such an approach provides several major advantages when compared to the classical, statistically-based pattern recognition systems. Firstly, it classifies as an astronomer would, thus giving credibility to its conclusions. Secondly, it provides a natural avenue for machine discovery of new classes of objects thereby meeting one of the major goals of CTI - serendipitous discovery. Thirdly, once set up, it will not make enormous demands upon the user's time.

The use of optical fibres in astronomical instrumentation has been going on at the Anglo-Australian Observatory (AAO) for the past four years. The major application of fibres has been for multi object spectroscopy, and a system known as FOCAP (Fibre Optic Coupled Aperture Plate) has been in regular use on the 3.9m Anglo-Australian Telescope (AAT) as a general user instrument for the past three years. FOCAP has undergone a program of continual development with improvements and new facilities to cope with the steadily increasing and diversifing demand. This paper seeks to describe the current status of the present FOCAP system and detail the experience we have gained from regular fibre observing. A brief survey of suitable astronomical fibres currently avaliable will be given as well as an outline of some future developments with fibres at the AAO.

Optical fibers used for the transmission of digital data offer the advantage of high speed and complete immunity from electromagnetic interference. They are particularly useful in astronomy for isolating sensitive detecting apparatus from noisy digital computers. A serial data link is described using relatively inexpensive, commercially available components (transmitter, receiver, cable and connectors) with an encoder/decoder circuit built using standard TTL digital logic components. The circuit encodes data into a self-clocked serial data stream in a modified NRZ (non-return to zero) format, allowing efficient data transmission. A decoder circuit is described for extracting the embedded clock and decoding the serial stream at speeds of up to 40 megabaud. Circuitry is presented for the operation of fiber optic receivers that do not operate well at lower frequencies, allowing for data transmission rates down to DC. Although designed for high speed fiber optic operation, the circuit is equally well suited to medium or low speed operation over wires and offers the advantage of not having to transmit a separate clock. Finally, a data error rate is measured by operating the link for a long period of time.

One limitation of fiber optic bundles presently in use for relaying optical images from plane to plane is that the smallest resolvable area in the image plane is determined by the diameter of an individual fiber. The fiber optic relay system described here samples the wavefront radiated from the image to be relayed so that resolution is limited only by wavelength and the diameter of the entire fiber bundle at its ends. Consequently, the system can be used to relay or to combine images in a diffraction limited system as, for example, in a large multiple mirror telescope. The relay assembly is made up of a fiber bundle whose opposite ends are arranged to form concave spherical surfaces. The image to be relayed is positioned at the center of curvature of the first spherical surface and the resultant image appears at the center of curvature of the second sphere. The end faces of all fibers occupy corresponding locations on the two spherical surfaces. The fiber bundle consists of approximately 7000 single mode optical fibers whose optical paths have each been adjusted equal to NA t 1/20A, where A is the wavelength of in-terest and N is less than 10. Total fiber area is approximately 10 percent of the available spherical area so that wavefronts from a number of bundles may be combined. Spacing of fiber ends is irregular in order to eliminate any grating-like effects. Aberrations of this system are not easily described in the classical sense. The f/no can approach 0.5 but the T/no is always much slower, because the number of fibers used does not fill the available area. Diffraction ray trace data is presented for a specific relay system which models the output from a Michelson stellar interferometer.

An automated fibre optic coupler (AUTOFIB) for multi-object spectroscopy has been design-ed and is currently being built for the Anglo-Australian Telescope (AAT) as a collaborative project between Durham University and the Anglo-Australian Observatory (AAO). The device consists of a high-speed, precision, fibre positioner and an interchangeable fibre module. The positioner places the fibre probes individually under microprocessor control and can set up a field of 64 fibres in less than 10 minutes.

Multi-object fibre optic feeds are being prepared for use with the Cassegrain spectrographs of the La Palma Observatory 2.5m Isaac Newton Telescope and 4.2m William Herschel Telescope. The techniques used for terminating and preparing the fibres are described. In order to maintain high efficiency when feeding f/15 and f/11 spectrographs with fibres, microlenses have been used. The preparation and setting up of a microlensed fibre system to feed the f/15 spectrograph of the INT from the f/3.3 prime focus is described. The performance of this system is demonstrated in the laboratory and details are given of the system under construction for the WHT.

A two-star photoelectric photometer has been designed for use on the 18-inch reflector at Appalachian State University's Dark Sky Observatory. The system will be composed of commercially available components, including integrated detector-amplifiers, stepper-motor controlled x-y stages, and automatic counters. Fiber optics will be used to route the light from the image plane to the detectors. Separate channels will monitor the sky beside each star.

The University of Calgary is working in conjunction with DAO (Dominion Astrophysical Observatory) and NRCC (National Research Council of Canada) to develop an experiment capable of performing UV (200-300 nm) spectroscopy on meteors. In order to perform UV spectroscopy on meteors successfully. The experiment must operate outside the earth's atmosphere because atmospheric ozone strongly attenuates UV shorter than 300 nm. NASA's GAS (Get Away Special) program has provided the University with an invaluable opportunity to place its experiment on a future shuttle flight. They will provide a cannister (5 cubic feet) in which to enclose the experiment. The GAS cannister will be attached to the inside of the shuttle's cargo bay. The spectroscope that will be used in the experiment will consist of an f/1.0 4.6 cm spectrographic camera (initially used on the 72-inch telescope at DAO), a Cs2Te MCP image intensifier and a CCD camera. The spectroscope will have a spectral resolution of 1 nm per pixel and a field of view of 13° x 10°. By studying meteors in the UV, it will be possible to detect the spectral lines of elements and compounds that have lines in the visual and near IR region coinciding with very strong fe and Ca lines. In the mid-uv, there exists an "iron-window", an under abundance of strong Fe lines. This window will permit the detection of Be, Zn and possibly B, all of which were previously undetected in meteor spectra. Thus, a comprehensive survey of meteor composition, derived from the meteors' UV spectra, will supplement previously existing data in the visual and near IR. These data will provide vital clues as to the origins of meteors.

We are using a charge-coupled device (CCD) in a scanning mode to find new asteroids and recover known asteroids and comet nuclei. Current scientific programs include recovery of asteroids and comet nuclei requested by the Minor Planet Center (MPC), discovery of new asteroids in the main belt and of unusual orbital types, and follow-up astrometry of selected new asteroids we discover. The routine "six sigma" limiting visual magnitude is 19.6 and slightly more than a square degree is scanned three times every 90 minutes of observing time during the fortnight centered on new moon. Semiautomatic software for detection of moving objects is in routine use; angular speeds as low as 11.0 arcseconds per hour have been distinguished from the effects of the Earth's atmosphere on the field of view. A typical set of three 29-minute scans near the opposition point along the ecliptic typically nets at least 5 new main-belt asteroids down to magnitude 19.6, but we do not follow all of those. In 18 observing runs (months) we have recovered 43 asteroids, discovered and reported astrometric and photometric data on 59 new asteroids, consolidated 10 new asteroids with orbital elements, and reported photometry and positions of 22 comets. We outline our future scientific programs that will take advantage of the expected performance of a Tektronix TK2048M-011 "thick", "quiet", cosmetically clean, front-illuminated CCD with 2048 x 2048 pixels that we have ordered.

The search for, the detection of, and the discrimination among moving objects has been a labor-intensive, time-consuming task. Earth-approaching asteroids, tenth planets, comets, meteoroids, and so on fill part of the solar system in an unpredictable way. Using state-of-the-art, low-light-level, beam-scanned, electron-bombarded silicon television cameras on a telescope equipped with a microprocessor/minicomputer control system, we have searched for Earth-approaching minor planets for several years. Two gener.tions of ever increasingly sophisticated analog and digital video data handling and image processing have now been upgraded with the addition of a VICOM Systems Inc. model 1800 digital image processor. The VICOM, with its own microprocessor, real-time digital video disc storage, and other peripherals allows for the real-time implementation of video processing algorithms which discriminates moving objects from the stellar background. As of the submission date of this abstract, all this is working at our observatory in a semi-automatic mode. We hope to have this fully automatic with the replacement of the current cameras by state-of-the-art, very low noise, very high transfer efficiency, and high quantum efficiency charge coupled device cameras. The capabilities of this system would be such that searches for variable intensity (but fixed in position) sources are feasible too. Whether supernovae, novae, flare stars, or other types of variability are sought, this would be the next step for our observatory. However, the detector/data handling/control/image processing problems associated with this task are much more difficult to solve.

The ARC 3.5m observatory is being designed to permit rapid switching among seven standard instruments and to encourage routine use by remote observers. Remote operation of the telescope and instruments will be required over short distances on-site as well as by astronomers at their home institutions, since the control room and laboratory facilities will be located in a separate building to minimize thermal effects on seeing within the telescope enclosure. In this paper, we discuss our design goals, the proposed architecture for the system, and our progress in development of software to implement the user/telescope and user/instrument interface. Prototype software already in operation includes a multi-window, event-driven, graphically-oriented environment implemented on a Macintosh computer for operation of 32-channel photometer and spectrometer systems. Communications between the Macintosh and a Multibus based instrument control computer occur over a single serial link which can include telephone lines. Controls are implemented graphically via pull-down menus and "control panels" which can be easily tailored to accommodate requirements of different instruments. Status information and data are returned from the instrument control computer and displayed continuously in real time. Among the data display "windows" are multichannel (1-32 channel) strip chart displays and integrated spectra which include statistical errors of individual data points. The software is written in an object-oriented language based on FORTH. Using this approach, a very sophisticated, reliable user interface can be implemented by a single programmer in a short time (several weeks). A similar con-trol environment which will include both telescope and instruments will be implemented via communications between a Macintosh or similar "smart terminal" and a single observatory computer which communicates with additional telescope and instrument control computers.

The importance of several unique features of the design of the Mark II camera has been proven in use over the past few years. The camera is now completely operational, and the maintenance laboratory has been moved to the observatory at Mount Locke. Recent improvements include: better baking; a new o-ring preparation utilizing high vacuum baking, aluminization, and oxidation; a new rough pumping system that eliminates waiting for pumpdown and reduces the risk of damage due to leaks and user error; a complete users manual; and a cassette magazine for the film. Photocathode lifetimes are now very long, and the photometric and astrometric properties of the camera have been documented. The reliability of the camera is now unsurpassed by any other instrument at McDonald. New projects include: the construction of an already designed motorized tracking guider; completion of the mechanical design and construction of a focal reducer for the 2.1m telescope; the production of K2CsSb(0) photocathodes to increase sensitivity and spectral range; the construction of a new film outgassing bench to allow large numbers of plates to be taken in one night; the installation of a CCD at the electronic focus of the Mark III camera; and the construction of the already designed 9cm., automatic, Mark IV camera.

The introduction of large area arrays provides new technical challenges in the retro-fit of existing astronomical instrument dewars and in the design of new dewars. The evolutionary implementation of cryogenically cooled dewars for visible CCDs at NOAO is described. The performance specifications required by applications in ground-based astronomy are discussed. Various dewar designs are presented and performance characteristics are given where applicable.

A photometer has been designed and built for the study of rapid photometric variations in a variety of astronomical objects. The photometer uses a CCD in its image storage mode to achieve relatively rapid readouts and to provide continuous, simultaneous monitoring of multiple stars and sky.

The ESA Photon Counting Detector, a scientific model for the Faint Object Camera of the Hubble Space Telescope, has a time-resolved imaging mode in which photon-counts are recorded separately for every frame (normally 30 msec long) and for every pixel (a 512x512 format is normally used). The system and its operation at the telescope are described, as well as some of the data reduction facilities. A discussion and sample observations are given for astronomical applications such as fast photometry of known sources, search for optical counterparts of variable sources, and image sharpening.

A new CCD-based image photon counting system (IPCS) has been developed at University College London as a successor to the original Plumbicon TV based IPCS that has been used extensively at the Anglo-Australian Telescope, the Isaac Newton Telescope (La Palma), Mt.Palomar and ESO. An engineered version of the new system is currently being designed, in collaboration with the Royal Greenwich Observatory, for use on the new 4.2m William Herschel Telescope on La Palma. The prototype of this new IPCS consists of a 4-stage EMI intensifier coupled, via a high speed, custom-built lens, to an RCA thinned CCD. Photon events detected by the CCD are then centred, using "interpolative centroiding" techniques, to an accuracy of 1/8th of a CCD pixel thereby attaining high resolution. Centroided data are then accumulated in a large computer memory. Included in this new IPCS are a scanning system to even out the effects of intensifier granularity and inaccuracy of the centroiding algorithm, a frame store to remove the CCD background and any double counting, and a correction for the charge transfer inefficiency of the CCD. Also included are CRT displays for showing both the incoming signal and the integrated picture. This paper describes the new IPCS and presents details of laboratory performance tests. Such tests have demonstrated high resolution and good detection efficiency in counting incoming photons. Also included are details on future developments of this system, including the possible use of a MCP intensifier (under development) and fibre optic coupling, as well as details of expected astronomical performance.

We describe the implementation of a panoramic photon-counting detector system of the Ranicon (resistive anode) type which is in operation at Mont Megantic Observatory. Particular emphasis is placed upon the modularity and flexibility of the system and the richness of its software environment. Spectroscopic applications now in course at Megantic are briefly described and some of the instrumental applications in preparation are mentioned.

The Hubble Space Telescope project, a joint project between NASA and ESA, will provide the European astronomical community with a source of high quality astronomical data on a routine basis. Recognizing the fact that data analysis facilities at many astronomical institutions in Europe are not adequate to analyze HST data in an efficient manner, ESA has, in cooperation with the European Southern Observatory, established the Space Telescope European Coordinating Facility (ST-ECF), with the explicit task (among others) of improving the data analysis situation. The European data analysis software situation is characterized by organizational and technological incompatibilities. This paper describes the concepts which were developed together with other European centers and with the Space Telescope Science Institute (ST ScI), to maintain compatibility on various differenet levels: data, applications programs, and command language. Standard Interfaces have been developed and are being used to port application S/W between systems. To make maximum use of the available resources it is attractive to implement a computer network, through which European astronomical institutions can communicate among each other and with the ST ScI. Our concepts and the results of initial trials are described. Perhaps the most important single development within the ST-ECF is the European HST Science Data Archive. This is a joint project with the ST ScI and covers the management and transfer to Europe of HST science data. The archive is based on dedicated Data Base Manage-ment hardware, and an optical disk data storage facility. The paper describes the implementation of the archive in a VAX/VMS environment and the interfaces to the data analysis system.

This paper critically examines the current software development scenario, in the academic as well as in the industrial environment. We identify areas where improvements should be made, and suggest possible techniques which can be used. The use of knowledge-based systems is considered, and their application to software development and to scientific research is examined.

An offset guider system has been developed for the 2.1 m S.P.M. UNAM Telescope. The system is based on an EEV P4310 CCD camera coupled to an image intensifier. A movable mirror sends a 4' x 4' field to the fixed detector system. It is done by means of collimating optics. The movable mirror scans the total field of the telescope, locating the object under study and then searching for an object suitable for guiding. The same system is used for the f/7.5, f/13.5 and f/27 ratios. The movable mirror and the collimating optics are controlled by a microprocessor system. This subsystem displays the movable mirror's field on a T.V. monitor and, if necessary, the field diaphragm of the mounted instrument. The same system allows the measuring of extrafocal images. A special algorithm determines the tilt and the decentering of the secondary mirror with respect the primary thus making the optical alignment routine easier.

As the data handling tasks required of space-borne astronomical instrumentation grow in size and complexity, more and more autonomy will be required of on-board control and analysis software. One area in which this is especially relevant is in the area of target acquisition. One simple method of target acquisition involves using on-board software to construct and downlink map images of areas of the sky as seen by on-board detectors. These maps are analyzed on the ground, and the on-board instrumentation package is then commanded from the ground to point in the proper direction. While this method is workable, it is far from optimal, since increasingly expensive real-time downlink data paths and uplink command paths are required. A more desirable method of target acquisition requires no real-time data or command links to or from the ground, and is carried out completely autonomously by on-board digital image processing and spacecraft pointing software. Such a software package was designed and implemented as part of the Flight Software for the High Resolution Spectrograph (HRS) for the Space Telescope (ST). This paper discusses the generic target acquisition process by using the HRS process as an example. Precisely because of the autonomous actions of the software, the ground-testing of autonomous on-board target acquisition software prior to launch is a crucial activity. In the absence of ST end-to-end target acquisition test facilities, a realistic test bed was nonetheless devised in order to test all features of the HRS target acquisition flight software. This paper briefly discusses some of the actual testing which has taken place. The HRS contains detectors which are 1-dimensional and provide a linear uniform lx500 pixel format, but sophisticated 2nd generation ST instruments are planned which contain 2-dimensional detectors providing formats of 2048x2048 pixels. Some implications of using 2-D detectors for target acquisition are briefly examined.

A high precision star tracker has been developed as part of the attitude measurement and control system of the German Roentgen-Satellit ("ROSAT"). A Peltier cooled CCD array in the focal plane of an 82 mm refractive lens is used to measure the position of up to 3 stars within a field of view (FOV) of 5.9° x 4.4°. The position of a fourth reference star provided by the telescope of the satellite as an alignment control can be monitored simultaneously. The CCD readout data are reduced by a fast preprocessor; subsequently a microprocessor com-putes the exact star position by pixel interpolation using a centroiding algorithm. Optical performance tests on the engineering model have demonstrated the capability to track stars of my = +6.5 with an accuracy of 2 arcsec and a noise equivalent angle of less than 1 arcsec with an update period of 1 sec. In scan mode the stars can be tracked with an accuracy of less than 10 arcsec at a scan rate of 5 arcmin/sec. The environmental tests of the system were completed in October 1985.

As part of the development of a Solar Oscillations Imager (SOI) Experiment for the pro-posed NASA/ESA Solar and Heliospheric Observatory (SOHO) mission a one-megapixel image acquisition and processing system has recently been installed at the Mt. Wilson Observa-tory. This system combines a high-speed CCD camera similar in design to that developed at JPL for the Solar Optical Telescope (SOT) program with a floating-point array processor and a large capacity disk storage subsystem. The CCD camera employs a 1024 x 1024 pixel virtual phase chip manufactured by Texas Instruments. The read-out rate of the camera is approximately 800,000 pixels per second, which means that the entire megapixel array can be read out in slightly more than one second. The JPL-designed camera is interfaced via a custom-design input/output port to a CSPI Mini-MAP floating-point array processor. This array processor currently contains 9 megabytes of memory and hence it can hold two 32-bit floating-point megapixel images at one time in addition to one 8-bit fixed-point image buffer. The array processor is employed to integrate several successive exposures from the CCD camera and to difference selected pairs of these integrated images. The processor can then normalize and calibrate the difference images in real time prior to outputting final images to disk memory at a rate of one calibrated difference picture every minute. The array processor can currently transfer these output images to two 704-megabyte disk drives, which provide sufficient storage capacity to hold one full day's worth (12 hours) of difference images. The stored images can then be archived on tape at 6250 BPI. The entire system is currently employed to record and difference narrowband images of the sun. In this way we can measure the Doppler velocities of the entire visible solar hemisphere once each minute at a spatial resolution of 2 arcseconds. We will describe this system here and show a sample of the images it produces.

The UV prime spectrograph at the Canada-France-Hawaii Telescope is the first instrument to be designed with an aspherized diffraction grating. This technique leads to all reflective Schmidt designs with a very small amount of optical surface on fast aperture ratio. A thin backside illuminated RCA CCD is now used as the detector. Since the detector is at the focus of an f/1 mounting, within the optical path, a minicryostat (5cmx5cmx3cm) was designed to minimize the central obscuration. This paper describes this new instrument and its performances.

The faint object spectrograph (FOS) is a highly efficient, fixed format spectrograph using a CCD detector and aimed at low resolution (15-20Å fwhm) spectrophotometry over the range 4000 to 10500K. The overall efficiency including atmosphere and telescope absorption is 12% at 7000K and a zenith distance of 27.5. It has a fixed spectrum format and is provided with an extensive set of on-line data reduction and analysis operations.

A multislit unit is being constructed for use with spectrographs on the 2.5m telescope at the Observatorio del Roque de los Muchachos. It will make possible the simultaneous observation of up to 12 spectra. The complete unit will fit into the available space on the telescope as a replacement for the conventional slit assembly. The area of the focal plane over which the slits move is 40x3Omm, corresponding to 4x3 arc min at the f/15 cassegrain focus. Each slit has length 2mm and can be set to within 18 microns of the required position. The slits are moved in the dispersion direction by wires pulled by linear actuators. The position of a slit is encoded from the voltage on an accurate potentiometer connected to the wire. Control is by a 6809 microprocessor which continuously monitors the position of each slit and adjusts it if necessary. The telescope main instrument computer will be used to translate object coordinates into encoded positions.

We have developed a versatile echelle spectrograph capable of moderate to high resolution precision spectroscopy. This highly stable instrument is in a temperature and humidity controlled room and is coupled to our 1.6 meter telescope via an optical fiber. Moderate resolution (λ/Δλ = 10000) spectra with near total spectral coverage from 400-900 nm are done at a Quasi-Littrow focus with no anamorphic magnification. A 200 mm f/2 camera focuses thirty-four orders of prism cross dispersed spectra onto a RCA SID501 CCD. A simple grating tilt can switch the beam to a cross-dispersing grating and long focus cameras allowing resolutions up to 80000. The high resolution beam has a more typical geometry with anamorphic magnification greater than one. Either small fibers or an image slicer can be used to obtain the highest resolution spectra. Observations conducted since Winter 1985 have shown the system to have excellent photometric and radial velocity characteristics; it is especially suited to long term, high signal-to-noise monitoring programs.

The echelle spectrographs for the Anglo-Australian Telescope (AAT) and for the William Herschel Telescope (WHT) are being developed following the same basic design, which in many respects, differs from previous echelle instruments. The purpose of this paper is to document some of the more innovative concepts for the benefit of future designers and prospective observers. The instruments were originally conceived around the idea of matching the spectrum format to the features of a particular detector - the University College London Image Photon Counting System (UCL-IPCS). Since construction started, very large CCD's have appeared on the horizon and this has called for certain modifications which are also discussed. Attention is drawn to various tactics for maximising throughput, including the new concept of a variable beam size. The arguments in favor of very large fused silica prisms for cross dispersion are summarised. The echellogram will be displaced in both dispersion directions on the detector and the method of accomplishing this is outlined. A novel automatic check of collimation is also described, together with an overview of important spectrograph functions such as those in the pre-slit area.

The MX Spectrometer is a remotely controlled multiple object fiber optic spectrometer head. Mobile fiber probes provide the capacity to obtain simultaneous spectra of many objects. Our experience with the Medusa aperture plate fiber optic spectrograph led us to design and build the MX with automated fiber positioning in the telescope focal plane. 32 stepper motor driven probes in a fishermen-around-the-pond arrangement position 64 fibers in the 45 arcminute field of the Steward Observatory 2.3m telescope. An onboard Z-80 microprocessor interfaces to 64 intelligent stepper motor controllers. The intelligent controllers allow simultaneous motion of all the probes for rapid field alignment. All fibers can be moved from one target pattern to another in less than 90 seconds. Two arcsecond diameter fiber apertures can be moved in steps as small as 0.1 arcseconds (10 microns). Probe collisions are prevented by software which maps the footprint of each probe in the focal plane. A failsafe electronic hardware system stops all motion in the unlikely event that two probes touch or encounter a foreign object. A second Z-80 with floppy disk drives is stationed in the control room to interface with the observer. This support microprocessor provides mass storage for coordinate lists as well as serial communications with the instrument microprocessor. We describe here both construction and actual operation of the MX Spectrometer at the telescope. The ability to record spectra of 32 objects plus additional sky spectra will greatly enhance our ability to study clustered systems of stars or galaxies.

A multiaperture spectroscopy device has been installed on the focal reducer at the Cassegrain focus of the Canada-France-Hawaii Telescope. The mask processing and ,positioning operations are described. Emphasis is placed on the speed and reliability of a fully automatic method controlled by a computer routine. Spectra obtained during the first observations at C.F.H.T. are shown as illustration.

The existing Risley prism at the 4 meter telescope is described. This device compensates atmospheric dispersion up to 60 degrees from zenith for the visible range. Under converging light at f/8 and f/30 the axial performance is good and the field imagery is mostly dictated by the properties of the primary-secondary configuration.

Using two spectrograph camera layouts, the instrumental constraints that influence their performance are explored. If the spectrograph already exists, at the secondary focus of a telescope, overall instrument length and detector implementation are the main impediments for camera future development. Ruling out a total instrument remake, trade-offs are suggested that may help during a transitional period.

The ESO New Technology Telescopy (NTT) (1),(2) to be operative by the end of 1988, will have two Nasmyth foci. The telescope diameter is 3.5 m and the Nasmyth F-no is F/11, corresponding to a scale of 187 um/arosec. At one focus the Infrared Spectrograph (IRSPEC) will be installed, while the other focus is reserved for an instrument operating in the 3000 - 10000 Å region.

We present the preliminary design for the Space Telescope Imaging Spectrograph (STIS), which has been selected by NASA for definition study for future flight as a second generation instrument on the Hubble Space Telescope (HST). STIS is a 2-dimensional spectrograph that will operate from 1050A-11000A at the limiting HST resolution of 0.05 arcsec FWHM, with spectral resolutions of 100, 1200, 20,000, and 100,000 and a maximum field-of-view of 50X50 arcsec. Its basic operating modes include echelle mode (like IUE), long slit mode (for simultaneous spectra at 1000 positions on the target), slitless spectrograph mode, coronographic spectroscopy, photon time-tagging and direct imaging. Photon counting MAMA detectors are used in the UV, while large format cooled CCDs are the visible and near-IR detectors. Research objectives are active galactic nuclei, the intergalactic medium, global properties of galaxies, the origin of stellar systems, stellar spectral variability, and spectrographic mapping of solar system processes.

A high-efficiency, extreme ultraviolet (EUV) imaging spectrometer has been constructed and tested. The spectrometer employs a concave toroidal grating illuminated at normal incidence in a Rowland circle mounting and has only one reflecting surface. The toroidal grating has been fabricated by a new technique employing an elastically deformable submaster grating which is replicated in a spherical form and then mechanically distorted to produce the desired aspect ratio of the toroidal surface for stigmatic imaging over the selected wavelength range. The fixed toroidal grating used in the spectrometer is then replicated from this surface. Photographic tests and initial photo-electric tests with a two-dimensional, pulse-counting detector system have verified the image quality of the toroidal grating at wavelengths near 600 Å. The basic designs of two instruments employing the spectrometer for astrophysical investigations in space are described, namely, a high-resolution EUV spectroheliometer for studies of the solar chromo-phere, transition region, and corona; and an EUV spectroscopic telescope for studies of non-solar objects.

Although previous work has emphasized the usefulness of gratings placed at grazing incidence in converging beams1'2 for extreme ultraviolet astronomy specific applications may well require a design that can be placed behind a slit at the prime focus of a telescope. Hence we have designed a grazing incidence spectrometer that accepts a diverging beam and performs moderate resolution spectroscopy. To enable a comparison with other recently pro-posed designs we limit our analysis here to an extreme ultraviolet spectrometer covering the 100 to 350 A bandpass. We assume that the diverging beam from the telescope has a focal ratio of 10. We find that an efficient spectrograph can be designed which has as its first optic an off-axis ellipsoid used at grazing incidence to refocus the diverging beam and increase the focal ratio. Any of a variety of gratings could be placed at grazing incidence in this slowly converging beam; we study in detail the performance of a plane, straight groove variable line space grating. Grating aberrations are evaluated analytically and verified by geometrical ray-tracing. Off-axis aberrations are determined by geometrical raytracing. The resolution of the system is a function of the telescope image quality. An optimum design is found by varying six parameters and maximizing the average resolution over the bandpass, assuming a telescope blur of 75 microns (corresponding to 1.5" at the focus of a telescope with focal length 10 meters). The resolution of this optimized design varies from 300 to 800 across the bandpass. The spectrometer can be located far from the optical axis of the telescope and fed by a small grazing incidence pick-off mirror. This represents a spectrometer which can be removed from the light path easily and thus would be a particularly compatible adjunct to other spectrometers that also utilize the f/10 diverging beam.

We have considered various designs for a high resolution spaceborne spectrometer for point-source astronomy in the 400 - 1200Å wavelength region. Our designs utilize as the pri-mary collecting optic a 1-meter normal incidence primary mirror of 1"quality and are constrained to fit within an envelope defined by the size of the advanced Spartan space platform now under consideration for development by West Germany. We find the most efficient design to be a multiple Rowland circle spectrograph, with four toroidal gratings each intercepting a fraction of the beam from the primary mirror. The advantage of this design is that each spectrometer can accept a relatively slowly diverging beam (thus reducing grating aberra-tions) without the loss of efficiency caused by an additional reflection or the magnification of the primary mirror blur and pointing jitter that would be introduced if a secondary mirror were used to slow the beam. We examine the detector requirements for the multiple Rowland circle spectrometer and find that no appreciable loss of resolution occurs if the circular tangential focal surface is approximated by two flat detectors. Furthermore, each pair of flat two-dimensional detectors can receive the spectra from two of the toroidal gratings simultaneously, thus reducing the number of detectors and associated electronics required. The specific parameters of the design (line densities, detector size and pixel size , etc.) are dependent on the drift rate of the space platform. We present a design optimized for an uncorrected drift of This design delivers a resolution of 3000 - 4000 and an effective area (including detector quantum efficiency) of 1 to 6.5 square centimeters across the bandpass.

We have designed a grazing incidence spectrometer to measure the diffuse astronomical background between 80 and 650A. The instrument uses a wire grid collimator to define a 40' by 15° field of view. After passing through the collimator, radiation impinges on an array of plane reflection gratings. An array of mirrors then focuses the diffracted light through a thin filter onto a microchannel plate detector. The resulting system focuses to a few arc minutes and produces spectra with 10 to 20Å resolution. The instrument is currently undergoing integration for flight in 1986. Performance characteristics of the spectrometer are discussed.

We describe the development of a 25mm microchannel-plate position sensitive photomultiplier tube suited for a variety of applications in ultraviolet astronomy. Specifically, this detector is intended for the FAUST payload, to be flown on the EOM-1 Spacelab Mission in 1986, and is designed for wide field (8°) imaging surveys in the vacuum ultraviolet band between 1400A and 1800Å. The detector is a sealed device comprising a CsI photocathode deposited on the inner surface of a MgF2 window, a stack of microchannel plates, and a wedge and strip two dimen-sional position sensing anode. Photoelectrons emitted from the photocathode are collected and multiplied by the microchannel plates to give gains of the order of 107 electrons per event. The resultant electron cloud is incident on the anode where the charge is divided between the three electrodes (wedge, strip, and zigzag) in a fashion that is a linear function of the event location coordinates on the detector face. This device is usable over X1200Å to 2000Å and exhibits a peak quantum efficiency of >10%. The quantum efficiency varies by less than 25% over the field of view. Position reso-lution of the order of 100 p.m FWHM has been achieved and is maintained to 50,000 cps. Image nonlinearity of less than 1% is obtained with stability better than 1:500. The wedge-and-strip principle requires only three anode signals. The flight instrument electronics is therefore simple, comprising chiefly of three charge amplifiers and three analog-to-digital converters. In a space flight application, this simplicity is a benefit to the reliability, cost and schedule for payload fabrication, test, and integration. For these reasons, as well as its performance, detectors of this kind may find applications outside the FAUST program.

X I present a new design for a moderately high resolution (λ/Δλ = 103) nebular spec-AX. trometer incorporating a mechanical precollimator with an objective Echelle grating followed by a cross-dispersing Wadsworth configuration. The design uses the minimum number of reflections to maximize extreme ultraviolet reflectivity, and yields a relatively large field of view for optimal sensitivity to diffuse emission. The configuration is compact, uses conventionally fabricated optical components and existing microchannel plate detectors, yet achieves sensitivities near the limit imposed by Liouville's theorem.

The radiometric response function of the High Resolution Spectrograph (HRS) for the Hubble Space Telescope (HST) was measured in 1984. From the system efficiency measurements made in the HRS echelle modes, we have deduced the relative efficiencies ("ripple" functions) of 14 HRS echelle orders. The results agree well with a theoretical model of echelle efficiency, except for a residual shift of apparent blaze angle with spectral order. The discrepancy between nominal and fitted blaze angle decreases with increasing order number.

We present the results of a comparative analysis of Hettrick Bowyer Type II (HBII) and Wolter Schwarzschild Type II (WSII) optics for use with Rowland circle spectrometers. We find that the HBII can substitute, with few exceptions, for a WSII in any Rowland circle with little change in spectrometer performance or design. This is demonstrated by raytraces through both instruments. However, the HBII telescope offers several clear advantages over the WSII in these configurations. Because the HBII employs a virtual focus, it requires a much shorter instrument length than a WSII. For example, a 3-meter Rowland circle spectrometer, fed by an optimized, f/10, 1-meter diameter WSII has a total instrument length of 6 meters. If a HBII is used to feed the identical spectrometer, the entire instrument length can be as little as 3 meters. In addition, the improved imaging gained with the larger graze angles of the HBII design results in better resolution in slitless operation modes.

Santa Barbara Research Center has developed a 58 x 62 staring mosaic focal plane array (FPA) consisting of an indium antimonide detector array hybridized to a silicon direct read-out (DRO) multiplexer for use in infrared astronomy. The detectors are sensitive to radia-tion from 1 to 5 μm. Each detector in the array is connected to the DRO via an indium bump contact and has its own high-impedance, low-capacitance buffer amplifier. The requirements of infrared astronomy are described and compared to the performance of this FPA. These requirements include low dark-current and readout noise for near-BLIP performance and full-frame readout times from 32 ms to more than 20s. Other features of the array include a nondestructive readout capability and user-controlled sequencing of the DRO, i.e., random access. Both modeling and measurements of the FPA performance are presented. Results include responsivity, noise, NEP, linearity, and dark current. These FPAs are currently available to the astronomy community along with a set of drive electronics which operates the FPA and provides analog signal processing.

The development and evaluation of an integrated array of antimony-doped silicon detectors is described. The spectral range of extrinsic silicon integrated arrays useful for low-background infrared astronomical applications is extended to about 31 pm with this development. The 58 x 62-element array is accessed by a direct readout (DRO) multilplexer. The device is evaluated with a flexible microcomputer-based drive and readout electronics system in a low-background test dewar. Acceptance testing indicates single-pixel NEPs in -18 the mid-10 W/?Hz range, and good global uniformity statistics.

The design of a near infrared imaging system for a large ground-based telescope is described and discussed. Solid-state hybrid detector arrays are used. The performance modelling of such devices is considered and their potential for IR astronomy is discussed.

A technique, developed by Cincinnati Electronics and Jet Propulsion Laboratory, of utilizing charge integration with InSb photovoltaic arrays is described. This device uses a MOSFET switched silicon multiplexer to sequentially connect each detector to a common line and JFET amplifier. A noise model is presented which can be used to predict performance of the detector/multiplexer system. Performance prediction curves are included that allow comparisons and decisions related to system design. This noise model can be generalized to calculate performance characteristics, particularly D* and NEP, for any detector/multiplexer configuration.

Mercury cadmium telluride is a promising detector material for low background astronomical imaging and spectroscopy in the near infrared. We have begun a program to measure dark current as a function of temperature in both isolated HgCdTe diode arrays and arrays which have been bump-bonded to silicon multiplexers. In this paper, we describe measurements of 2.5 μm cutoff diode material and a hybrid array with a CCD readout. Measurements of 52 μm x 52 μm detectors indicate that id may be less than 50-100 e-/sec at 87 K.

Guidelines for specifying instrument magnifications and image quality are derived to aid in evaluating optical layouts that are suitable for a proposed 1-5 micron filter photometer using a 2D array. The instrument is required to have a wide range of discrete magnifications and to be usable at the Cassegrain focus of three Kitt Peak telescopes. Two optical layouts, both lens-mirror hybrids, are described and their optical properties are compared.

A high-speed digital sequencer has been developed for controlling the new generation of 58 x 62 pixel format IR imaging devices. The design features a 92-bit x 1K RAM control-store memory and runs at 10 MHz. The controller is unique in that 34 signals are made directly available to an external device for use in sequencing the clocking of the detector. A Forth program for assembling user-defined code sequences has been written to assist in programming the controller. The Forth cross assembler runs on a DEC PDP or LSI-11 computer and generates downloadable microcode which can be readily changed.

We have constructed an infrared imaging system that achieves diffraction-limited spatial resolution (about 0.8 arc seconds) at 10 um on 3 meter ground-based telescopes. The system uses a linear array of sensitive HgCdTe photodiodes, scanned in the direction perpendicular to the array axis, to form two-dimensional images. Scans are completed rapidly enough to freeze atmospheric fluctuations. Individual detectors are small compared to the diameter of the Airy disk, and images are oversampled heavily in the scan direction. This method has a number of advantages for studying small fields with very high spatial resolution, and has been applied successfully to the problem of directly imaging faint circumstellar dust shells.

We present an overview of the current plans for the development of optical and infrared instrumentation for the Keck Telescope. The physical and optical properties of the telescope are summarized. We also discuss the instrument module concept developed for rapid instrument change. Preliminary descriptions of the instruments currently under consideration are given.

A high resolving power mid-infrared, astronomical Fourier transform spectrometer has been developed for use at ground based telescopes. The spectrometer is designed to operate over a broad wavelength range from l0 to 500 Pm, (i.e. 1000 to 20 cm-1), the specific range being defined by the choice of beamsplitter, filters and detector. The spectrometer employs a folded optics configuration in which a double-sided mirror moves on a linear carriage to produce a factor of four multiplication of optical path difference. This results in a compact instrument in which an unapodised resolution of .01 cm-1 is achieved with a mirror translation of only 12.5 cm-1 and which can easily be mounted at the f/35 Cassegrain foci of infrared telescopes. A major feature of the system is the ability to compute, display and monitor any part of a 32,768 point spectrum of a phase corrected single-sided interferogram in real time. This is accomplished by efficient Fourier transform routines which exploit the inherent symmetry properties of the interferogram. These and other features of the complete spectrometer system are discussed.

Improved sensitivity to light and uniformity of response has been obtained using various surface treatments of thinned CCDs. Procedures are described and some results are shown. The Lick Observatory data acquisition system is briefly described. A new thermoelectrically-cooled CCD camera for telescope guiding has been designed and built. The camera is described and preliminary results at the telescope are given.

A double density (15 microns per pixel) RCA CCD has been tested for astronomical imaging and spectroscopy at the Canada-France-Hawaii Telescope. This particular sensor has been treated with an anti-reflection coating to enhance performance in the ultra-violet (UV). The CCD is compared to the 30 micron RCA chip that has been in operation at CFHT for some time. In many applications instrumentation and seeing quality on Mauna Kea demand better sampling than 30 microns. The RCA double density chip was purchased and evaluated with these particular constraints in mind. Although the read noise is not as low as with some other three-phase CCD's, the unique backside treatment and permanent UV enhancement provided by RCA do give some advantages over some other sensors currently used.

A high-efficiency spectrograph for research on faint astronomical objects has been successfully built and is in regular use on the Multiple-Mirror Telescope (MMT). It is equipped with a motorized multi-slit mechanism to allow simultaneous observations of several objects. The combination of high throughput, multiple object capability, and low-noise CCD detector has allowed routine work to be extended to objects of V magnitude 23 and fainter.

By having the ability to thin CCDs with precise control of thickness and surface quality astronomers will be able to optimize chips for specific applications. Traditional chemical etching is used to reduce subsurface damage of the silicon crystal lattice. We demonstrate a means of mechanically thinning a TI 800x800 CCD with an abrasive slurry of aluminum oxide. Using the same techniques we can replace the abrasives with a chemical solution to eliminate subsurface damage. We also demonstrate a technique of mounting the CCD which retains the high quality surface generated during thinning. This requires the backside of the chip to be bonded to a glass window which closely matches silicon's thermal expansion properties. Thinned CCDs require backside treatment to enhance blue and UV quantum efficiency. Two methods are discussed which may be effective with this mounting system. Due to silicon's high index of refraction in the visible spectral region up to 50 percent of the radiation incident onto a CCD can be lost to reflection. We briefly discuss the improvements to be gained from AR coating the backside.

The C81020E is an electrostatically-focused image tube with a fiber-optic window which uses a thinned SID504 CCD operated in electronombarded mode to detect accelerated photo-electrons. To evaluate this device, a system was assembled using a commercially available camera electronic chain, a frame grabber, and a microcomputer data acquisition system. Preliminary data on geometric and electronic linearity, dynamic range, and lifetime under electron bombardment indicate that the device could be useful in astronomical applications. It can be used at low light levels in lieu of conventional CCD's, and probably will be satisfactory as a photon counter in the event centroiding mode. The charge cloud produced by the primary appears to spread out over more than one pixel, precluding use of the device as a simple digicon.

Although current interest is focussed on the emergence of large CCDs from Tektronix, for many applications smaller devices based on the TV imaging format are still relevant. Several manufacturers produce CCDs of this size, and deciding which to use for a particular application is of some importance. Over the past three years, the general purpose CCD camera system developed at' the Royal Greenwich Observatory (RGO) has been used to test, optimise and compare various types of CCD with further work being carried out on a variety of telescopes. The results of this work are described, dealing not only with the obvious features such as noise and quantum efficiency, but the more subtle aspects such as low light level performance and temperature dependence. The astronomical relevance of the various factors is considered and suggestions made as to the suitability of each type of chip for various types of observation. Finally, the future of these TV format CCDs is considered in the light of the likely availability of much larger alternatives.

Preliminary findings are presented for a new approach that significantly improves the quantum efficiency of the current generation of high-performance, thinned, backside illuminated silicon CCDs. Experiments have shown that the application of an ultra-thin (less than 4a) layer of metal with high work function to the backside of the CCD can yield 100% internal quantum efficiency in the visible, UV, XUV and soft X-ray regions of the spectrum. Theory and solid state models describing the new technique (which we refer to as the CCD flash gate), plus a considerable amount of experimental data are discussed in this paper. Specific recommendations for use of the flash gate in present and future CODs are also reviewed.

We discuss the applicability of CCD detectors to astronomical extreme ultraviolet (EUV) spectroscopy ( 100 - 1250 Å). The advantages of CCDs in this spectral region include internal electron yield (~ 3-30 e-/per detected photon), the potential for very high quantum efficiency (~ 50-90 percent ), and broad wavelength response. Visible light suppression is achieved by a combination of low grating scattering, great er than unity electron yield in the EUV, and various filter techniques. For the current generation of (Cl),s detection of only a few EUV photons will rapidly overwhelm the read noise ( < 5-10 e- RMS). Thus for all practical S/N ratios used in ast-ronomical spectroscopy, read noise will be negligible compared to the poisson statistics of the detected photons. We will discuss a model based upon experimental data for the quantum efficiency and electron yield of CCDs in the EUV, outline charge trapping and transport considerations, and describe a program to obtain CCDs with stable EUV/UV quantum efficiency for windowless spaceflight operation.

The development of large-format CCD arrays at Tektronix (Blouke, et al., 1985) has led to the proposal for such an array to be placed at the focus of the Hubble Space Telescope, where it will potentially become the workhorse of space astronomy during the 1990s. The characteristics of the Tektronix CCD make it a near-ideal focal plane detector for space astronomy in the 100 - 1000nm wavelength range: the large format (2048 x 2048 pixels) and 27um pixel size are closely matched to the HST resolution and field of view for the radial-bay; the low readout noise (goal -3 electrons rms) and large full well capacity (750,000 electrons) lead to a dynamic range which is orders of magnitude greater than the first-generation HST cameras; the quantum efficiency and overall throughput are close to theoretical limiting values; and the excellent charge transfer efficiency ensures that the near-ideal sensitivity limit of HST can be reached. The Advanced Radial Camera will greatly expand the capabilities of the first generation instruments on HST. It will allow the continuation of all programs begun with the first generation HST cameras, while enabling their completion on significantly shortened timescales.

A summary is given of measurements of the photoelectron counting efficiency of a variety of image intensifiers manufactured by several different manufacturers. With one exception, the counting efficiency is disappointingly low. Typically only slightly over half of the photoelectrons from the first photocathode produce a detectable signal on the output phosphor screen in most diode-type intensifiers. For microchannel plate intensifiers, the counting efficiency is even less. Results are given of experiments intended to produce diode-type intensifiers having improved counting efficiency. Improvements have been achieved through refinements in the manufacturing steps of the phosphor screen, the tube component identified as responsible for the lost photoelectron pulses in diode-type tubes. The best tubes so far have counting efficiencies of 70%, which represent an improvement by a factor of about 1.4.

For the 4.2 m William Herschel Telescope (WHT) to be erected at the Roque de los Muchachos Observatory on La Palma in the Canary Islands, the Netherlands Foundation for Radio Astronomy has developed a conceptual prototype integrated CCD controller in cooperation with the Royal Greenwich Observatory. It has been demonstrated that preamplifier noise limited system performance of 5 rWviliz. can be obtained with a compact low power controller that uses switched mode power supplies throughout. This performance has been realized by applying a differential preamplifier with balanced outputs between the CCD in the cryostat and the correlated double sampler in the controller, realizing over 100 dB common mode rejection. A complete system to drive two independent heads simultaneously each with either a large CCD or an array of four small chips can be realized into a small crate using single height long Eurocard modules and dissipate about 50 W.

Adaptive Optics Associates, Inc. (AOA) has recently constructed a new commercial low-light imaging detector which we call the P4oton Digitizing Camera (PDC). Based on the PAPA detector developed at Harvard Universityl", the PDC generates (x,y,t) coordinate data of each detected photo-event. .Because the positional address computation is performed opti-cally, very high counting rates are achieved even at full spatial resolution. Careful optomechanical and electronic design results in a compact, rugged detector with superb per-formance. The PDC will be used for speckle imaging of astronomical sources and other astro-nomical and low-light applications.

The development of a new microchannel plate, photon counting detector with an analogue readout method based on a resistive anode is reported. This detector exhibits extremely high, stable electron gains of 108. At this gain, the spatial resolution is no longer primarily limited by the noise of the resistive anode so that digital methods of readout, such as discrete conductors, lose their advantage. These detectors can be readily scaled to 40mm and 70mm formats to match plate scales of 2 meter and larger telescopes. New, high speed digital electronics are described which replace older analogue methods and which fully exploit the high spatial and time resolution made possible by gains of this level. Analysis of the theoretical performance of this detector is performed and shows that the major limitation to the spatial resolution is the proximity focus of the photocathode and the first microchannel plate. The detector has been mated to an echelle spectrograph developed at the University of Michigan for installation on the new 2.4m facility at McGraw-Hill.

The High Resolution Spectrograph for the Hubble Space Telescope uses two pulse-counting Digicon detectors to record ultraviolet spectra in the wavelength range 1050 Å ( A< 3200 Å. One of the attractive features of the Digicon is the well behaved relationship between the measured count rate and the true photon arrival rate. At low rates the relationship is linear with unit slope and zero offset. At higher rates the intervals between photon arrivals become comparable to the response time of the electronics and not every pulse will be counted. This "paired-pulse" effect causes the relationship to depart from linearity at high count rates and must be corrected for during the reduction of raw data. In this paper we describe the characteristics of the pulse-counting circuits, and present two analytic equations which quantify the dead-time losses. During the ground-based calibration of the HRS observations of bright emission lines in the spectrum of an argon miniarc lamp were made to calibrate the non-linearity. By attenuating the light with neutral density filters the same spectrum could be recorded with input count rates ranging from 20 to 1.4 million counts per second per channel. The less attenuated spectra are severely distorted by the dead-time losses. Comparison with the undistorted low count rate data allows a detailed analysis to be made of the pulse counting characteristics over five decades of input event rates. The fitting formulas allow data with input rates up to 100000 counts per second to be linearized with no systematic errors. Our equations and calibration methods should be generally applicable to all kinds of multi-channel pulse counting detectors, which are becoming increasingly common in both ground and space based astronomical instruments.

As part of a collaborative project to develop a high gain microchannel plate (MCP) intensifier for large area photon counting detectors, a detailed study has been carried out on a prototype device. A calibrated absolute faint light source has been used for the DQE measurements, and direct comparison has been made with a number of photomultiplier tubes. As a result of this work it became apparent that an advanced production chamber would be required to make it possible to introduce modifications readily into the design, and to improve quality control during manufacture. Such a chamber has recently been commissioned. A proposal is discussed for a design of intensifier using a combination of curved and straight channelled MCPs.

We report on the capability and initial use of the MAMA detector in the time-tag mode. The detector hardware currently in use consists of a visible-light detector tube with a semi-transparent photocathode proximity-focused to a high-gain curved-channel MCP. The photoevents are detected by a (256 x 1024)-pixel coincidence-anode array with pixel dimensions of 25 x 25 microns 2 connected to charge-sensitive amplifiers and event-detection circuitry. In the time-tag mode, the detector delivers the pixel address and the time of arrival for each detected photon to an accuracy of 10 pS. Detector count rates of >2 x 105 counts s are possible in the time-tag mode, but the maximum count rate is limited at this time by the speed of our data acquisition hardware. The MAMA detector in the time-tag mode is currently being evaluated in programs of astrometry and speckle imaging.

We describe microchannel plate QVICP) detector systems that use wedge and strip anodes for imaging radiation in the range 50A to 8000Å. Test data, obtained from the calibration of the Extreme Ultraviolet Explorer Satellite (UVE)1 detectors and the AMA` EUV aeronomy sensor is presented. We have investigated the long term performance characteristics of Z stack microchannel plates and show that stable operation is achieved after an initial settling period. We discuss conventional and new anode designs and their potentiar increasing, the position resolution. Position resolution of better than 80 microns FWD has been attained for 50mm devices and is shown to be essentially independent of event location. Image nonlinearity of less than 0.3% has been achieved with stability to better than 0.1%. We demonstrate that differential nonlinearity of less than 1% and stability to better than 1% have been obtained. Measurements of the quantum efficiency of a CsI photocathode on an MCP are presented, demonstrating efficiencies as high as 47%.

The Johns Hopkins University is currently developing an instrument to fly aboard NASA's space shuttle as a SPARTAN payload in the late 1980's. This SPARTAN free flyer will obtain spatially resolved spectra of faint extended emission line objects in the wavelength range 750 -1150 Å at ~2 Å resolution. The targets selected for the first mission of the payload include the Io plasma torus around Jupiter and the Cygnus Loop supernova remnant. The use of two-dimensional photon counting detectors will give simultaneous coverage of the 400 A spectral range and the 9 arcminute spatial resolution along the spectrometer slit. The progress towards the flight detector is reported here with preliminary results from a laboratory breadboard detector, and a comparison with the one-dimensional detector developed for the Hopkins Ultraviolet Telescope (HUT) 3. A hardware digital centroiding algorithm has been successfully implemented. The system is ultimately capable of 15 pm resolution in two dimensions at the image plane and can han-dle continuous counting rates of up to 8000 counts s -1.

I describe and compare several image enhancement techniques, useful for morphological analysis of galaxian (or cometary) images. Such techniques can be used to search for, and investigate the properties of dust lanes, stellar disks or rings, jets, shells, tidal distortions, etc. Applications of the techniques are illustrated on CCD images of the peculiar galaxy Arp 230; this object has a rich morphology, indicative of a merger of two disk galaxies.

Evaluated are four different centroiding algorithms for use in the Kitt Peak National Observatory Photon Counting Array (KPCA). Computer simulations are performed on each of the four different methods. Variations are applied in the simulations and a decision is formed resulting in the method configured in hardware. The algorithm needs to control the placement of the center of a photon splash to 1/8 + 1/8 of an array element. The overall photon splash will acquire an approximate spatial area of 5 x 5 array elements. Restricted by the need to centroid in real-time, the algorithm must handle incoming array data at a rate of 20MHz. Both 3 x 3 and 5 x 5 spatial algorithms are considered. Simulation results of the selected algorithm are confirmed in the centroider hardware, giving the resolution required. Based on the simulations evaluated, new real-time configurations are being con-sidered for future implementation.

This paper presents an algorithm for the optimal estimation of the centers of digitized stellar and semi-stellar images. If the image profile function is known, this an iterative, interpolative algorithm provides an optimal estimation of the image center. The algorithm converges rapidly and produces answers identical to those obtained from a X² -fit of the assumed profile function to the data.

Instrumental efforts in Multiple Aperture Fiber Optic Spectroscopy (MAFOS) at Steward Observatory have greatly increased the observing efficiency for spectroscopy of clustered objects. The Medusa aperture plate spectrograph and its successor the MX mobile fiber spectrometer allow simultaneous spectra of 30 or more objects to be recorded. To carry out an effective scientific program, this increase in data collection rate must be matched with similar speed increases in data reduction over conventional techniques. We have developed the Medusa Automatic Reduction System (MARS) to reduce intensified CCD spectra from Medusa and "barefoot" CCD spectra from MX. Data produced by these multiobject spectrometers is in the form of two-dimensional images containing 44 to 66 individual spectra. The philosophy of MARS is to make "hands off" reduction possible for large quantities of similar data, and to accomplish this task in time comparable to that required for observing. Both hardware and software improvements were required to extract one-dimensional spectra from two-dimensional data frames in an efficient manner. A VICOM image processing system is used for image display and inspection. Two-dimensional processing of the CCD frames and extraction to one-dimensional files is carried out on a DG MV/10000. Our ridge finding algorithm for automatic location of the spectra in the CCD frame is described. Simultaneous extraction of calibration frames allows an automatic wavelength calibration routine to determine dispersion curves. Line measurement and cross-correlation techniques are used to determine galaxy redshifts. Sky subtraction, flux calibration, and flat field techniques on multiple fiber systems are also discussed.

This paper describes the echelle reduction package available within MIDAS, the ESO image processing system which runs on VAX computers. The reduction method is currently used to process spectra from the echelle spectrograph at the Cassegrain focus of the ESO 3.6 m telescope at La Silla, Chile. The package is very modular, with intermediate data structures allowing the communication between the different reductions steps. As a consequence, the reduction method is flexible and can be adapted to a wide variety of instrumental configurations. The basic reduction steps: wavelength calibration, data extraction, and flux calibration are performed with minimal user interaction. Some results showing the accuracy of the method are presented.

Many image restoration methods have been developed over the years ranging from the simple-minded inverse matrix method to the hard-to-understand maximum likelihood methods. Authors have tended to use high-falutin' language (projection onto convex sets, regularization methods, solution of the Fredholm equation of the first kind, etc.) scaring off potential users. These users fall back on the Wiener filter and its many variants because it is understandable and relatively simple to program. I will simplify the approach to image restoration by presenting a class of recursive restoration algorithms based on the classical imaging equation. These algorithms were developed by employing simple algebraic identities to manipulate the imaging equation into recursive forms. Some of the algorithms naturally satisfy the positivity constraint, making them useful for superresolution of degraded imagery. These recursive techniques are simple to understand and to implement, and give results approaching those of the most sophisticated image restoration algorithms.

Two almost identical echelle spectrographs are under construction: for the 3.9m Anglo-Australian (AAT) and the 4.2m William Herschel (WHT) telescopes. The versatility of the optical design requires the simultaneous control of many mechanical functions, so a high level of automation under computer control is essential. The control and data reduction software packages will be highly interactive and will be implemented at the observatories and at home institutions, allowing prospective users to test the feasibility of proposed observing programs. Then, at the observatory, the accumulating data can be preliminarily reduced allowing real-time decisions to be made. The data reduction package should find applications to echelle data generally and the control system may easily be adapted to other instruments.

One of the major problems in dealing with the plotting of astronomical data is the accurate projection onto a two-dimensional surface of data based on a spherical coordinate system. AMP has solved this problem by using three map different projections - Mercator, Lambert conformal and polar stereographic.

The Image Reduction and Analysis Facility (IRAF) is a general purpose software system for the reduction and analysis of scientific data. The IRAF system provides a good selection of programs for general image processing and graphics applications, plus a large selection of programs for the reduction and analysis of optical astronomy data. The system also provides a complete modern scientific programming environment, making it straightforward for institutions using IRAF to add their own software to the system. Every effort has been made to make the system as portable and device independent as possible, so that the system may be used on a wide variety of host computers and operating systems with a wide variety of graphics and image display devices.

Tools for the reduction of long slit spectra within the Interactive Data Reduction and Analysis Facility (IRAF) at the National Optical Astronomy Observatory (NOAO) are described. The user interface (commands and special features) and the algorithms are discussed. Application of the reduction package to multi-slit images is briefly outlined. The author developed and supports the package at NOAO.

Three deconvolution techniques, CLEAN, Maximum Entropy Method (MEM), and L. B. Lucy's algorithm, are compared through both one-dimensional computer simulation and real astronomical data obtained at high angular resolution. These algorithms are evaluated according to speed, noise sensitivity, and resolution gain. For the simulations, we define the resolution gain needed to separate the components of an equal double as the ratio of the FWHM of the point spread function to the double separation. CLEAN is at least two orders of magnitude faster than the other two algorithms but is limited in usefulness to resolution gains near unity. Lucy's algorithm converges rapidly at first and achieves a resolution gain near two but requires smoothing techniques to reduce its sensitivity to noise. MEM is less sensitive to noise and is capable of resolution gains as high as four. These conclusions are also demonstrated with real astronomical data where MEM again produces the smoothest image consistent with the signal-to-noise.

A wavefront sensor has been designed to measure wavefront aberrations occurring during astronomical observations with high spatial and temporal resolution. Results of laboratory experimentation are presented. The sensor consists of a grating lateral shear heterodyne interferometer using a solid-state detector array. It has high light efficiency and is a self-referencing interferometer, thus allowing the use of white-light extended astronomical sources for the measurements. The wavefront sensor employs a heterodyne technique and is insensitive to intensity variations across the pupil and detector nonuniformities. To obtain a wavefront map, two sets of orthogonal wavefront difference data from the lateral shear interferometer are required. Both sets are measured simultaneously using a single laminar grating to produce the sheared diffraction orders and a single solid-state detector array placed in a plane conjugate to the telescope pupil plane. The signal from the detector array is digitized and fed into a computer where the wavefront differences are calculated. From these differences a wavefront map on a square array of 24 by 24 points is obtained by filtering the data in the spatial frequency domain. The time required for the acquisition of a complete set of data for one wavefront map is 13.3 ms. A wavefront reconstruction may be done within several milliseconds by using fast algorithms and floating point processor hardware.

We have obtained diffraction limited images for bright objects using various large astronomical telescopes, including the MMT, but the results have not been convincing for fainter objects. All speckle interferometric techniques utilizing image power spectrum or autocorrelation methods to obtain calibrated image amplitudes are subject to noise bias induced by the Poisson statistics of photon detection. This noise bias, independent of image frequency for ideal detectors with delta-function like point spread functions, must be removed to obtain calibrated image amplitudes. The noise bias characteristic of our intensified TV detector, both as a raster event localizer (faint object mode) and as a conventional image amplitude detector (bright object mode), is not only image frequency dependent but signal rate and signal distribution dependent as well. This precludes signal independent calibration of the characteristic bias functions and sets limits on object faintness for successful image recovery. Implications for detector development are noted.

Both video raster detectors and newly developed 2-D photoelectron event locators are now employed to collect data at large astronomical telescopes for diffraction limited optical imaging by a number of speckle interferometric techniques. The Knox-Thompson algorithm is among the best understood of these techniques. The image amplitude and phase integrations required for its implementation can be accomplished directly in event-coordinate space (in the extreme photon-limited case) more efficiently than in the more conventional Fourier transform space. We describe hardware to accomplish the integrations required for Knox-Thompson image reconstruction in real-time at the telescope, thus alleviating the principal data processing bottleneck in the present applications of diffraction limited astronomical imaging. Parallel digital recording of the photon coordinate lists permits subsequent off-line re-analysis of the data by other methods if desired. The basic system consists of hardware to localize discrete photoelectron responses in a CCD TV raster and for accumulation of the complex autocorrelogram. A generalization of the system can integrate triple correlations.

Multi-object slit spectroscopy with a wide-field telescope is potentially a very effective way of collecting large quantities of data for statistical analysis on a wide variety of astronomical object-classes. A pilot-model system with 40 channels has been developed for the 1.2-metre UK Schmidt telescope, and has been extensively tested with photographic detection of the spectra. The commissioning of a dedicated cooled CCD camera in the immediate future is expected to lead to the rapid implementation of a user service with the system.

Current techniques for infrared speckle interferometry treat all exposures with equal weight regardless of instantaneous atmospheric "seeing". In an effort to improve accuracy in measurements of visibility amplitude and phase, we utilize measurements of image quality to bin individual exposures. We have investigated several criteria including the second-order moment of the image (an estimate of the Fried parameter) and two sharpness functions considered previously in adaptive optics applications. Test results on bright stars indicate that the sharpness parameters are more sensitive to image quality. These tests also delineate the correlation between sharpness and the second-order moment and demonstrate the variations, sometimes periodic, in image quality caused by atmospheric seeing. Initial tests of a binning technique using real astronomical data exhibit improved accuracy and reduced sensitivity to atmospheric fluctuations.

The design and performance of the Imperial College hard-wired cross-correlator is described. Its use in conjunction with the Imperial College hard-wired autocorrelator is detailed and illustrated by a comparison of data reduced with and without cross-correlation.

A system to measure image size and motion has been designed, constructed and put into operation at the 4-meter Mayall telescope at Kitt Peak National Observatory (KPNO). The present system configuration consists of three elements. A 128 x 128 pixel CID camera is mounted at the south port of the Cassegrain instrument rotator. The CID camera is fed, under software control, by a movable mirror in the instrument rotator. Focusing is achieved by determining the point of zero-image motion when the f/8 beam is cut by a rotating Newall mask. Focal plane images are digitized and stored on digital disk for later determination of image size and motion as a function of time. The second element is a two-dimensional position sensor which records image motion induced along an air path located between the primary and secondary mirrors of the telescope. The third element monitors temperature of the telescope and air throughout the telescope structure. The system is described and preliminary results are presented.

In this paper, we describe an algorithm which permits recovery of a diffraction limited image from a set of turbulence degraded images. This algorithm, first proposed by Knox and Thompson, can be considered as an extension of the speckle interferometry technique. Turbulence dependence and photon noise limitations are considered and difficulties arising from phase dislocations in the Fourier transform are discussed. The effectiveness of the algorithm has been demonstrated for extended objects. Reconstructions for test patterns degraded in a laboratory turbulence simulation facility are presented. Requirements for a real-time processing system for present and future telescopes are discussed.

A photon-camera star tracker was developed for use in a ground-based optical stellar interferometer. A major advantage gained by using a photon camera as the light detecting element is that it allows the received image to be cross-correlated with any desired spatial function. A theoretical analysis is presented to show that proper selection of a cross-correlation function can result in improved tracking performance. In addition, the optical, electronic, and software design of the star tracker is discussed.