POINTS is a dual astrometric optical interferometer with nominal baseline length of 2 m and measurement accuracy of 5 microarcsecs for targets separated by about 90 degrees on the sky. If selected as the ASEPS-1 mission, it could perform a definite search for extra-solar planetary systems, either finding and characterizing a large number of them or showing that they are far less numerous than now believed. If selected as AIM, it could be a powerful new multidisciplinary research tool, opening new areas of astrophysical research and changing the nature of the questions being asked in some old areas. Based on a preliminary indication of the observational needs of the two missions, we find that a single POINTS mission lasting ten years would meet the science objectives of both ASEPS-1 and AIM. POINTS, which is small, agile, and mechanically simple, would be the first of a new class of powerful instruments in space and would prove the technology for the larger members of the class that are expected to follow. The instrument is designed around a metrology system that measures both the critical distances internal to the starlight interferometers and the angle between them. Rapid measurement leads to closure on the sky and the ability to detect and correct time-dependent measurement biases.

Newcomb is a design concept for an astrometric optical interferometer with nominal single-measurement accuracy of 100 microseconds of arc ((mu) as). In a 30-month mission life, it will make scientifically interesting measurements of O-star, RR Lyrae, and Cepheid distances, probe the dark matter in our Galaxy via parallax measurements of K giants in the disk, establish a reference grid with internal consistency better than 50 (mu) as, and lay groundwork for the larger optical interferometers that are expected to produce a profusion of scientific results during the next century. With an extended mission life, Newcomb could do a useful search for other planetary systems.

The accuracy of the relative metrology gauge developed for the proposed OSI and SONATA missions is improved to subpicometer level. An accuracy of 0.15 picometers is obtained in vacuum at time scales of a few minutes. A surface metrology gauge with an initial accuracy of (lambda) /1000 is under construction. Photometry accurate to better than 1 part in 10³ for the surface metrology gauge is demonstrated using a commercial grade, 8-bit, uncooled CCD camera and a commercial grade frame grabber at time scales of 10 seconds with a resolution of 320 by 240 pixels.

Recent advances in electronics and fast computer control allow to envisage extremely high spatial resolution observations of the Sun through the use of a compact array of phased telescopes. Several space missions (SUN/SIMURIS, SUN-SV, MUST/SIMURIS) have been proposed in that respect and will be briefly presented. Prospects for use of the space techniques for a solar array on ground are also indicated. Independently from the different mission concepts, solar interferometric imaging presents a unique case in the domain of optical aperture synthesis since the field- of-view is extended (larger than the diffraction spot of a telescope) and because the high resolution structures are evolving very rapidly and are naturally complex (low fringe visibility). These severe constraints drive solar arrays' design towards `compact' configurations (i.e. in which the spatial frequencies plane is filled) and real-time `cophasing' (direct-- hardware--zeroing of phase fluctuations by fine delay lines). They also influence the choice of the focal instrumentation which is optimum when using a subtractive double monochromator tunable over a large spectral range and providing narrow band filtergrams (up to 0.1 angstrom). We review the concepts and design issues of a solar interferometer and present numerical simulations and laboratory experiments of the system required to cophase an array of telescopes on a complex and extended field-of-view. Aperture configurations and image reconstruction are also discussed as well as the specific real-time metrology aspects of a ground array (atmospheric constraints derived from the performances evaluation of the ASSI Program).

The Navy Prototype Optical Interferometer (NPOI) is nearing the completion of the first phase of construction at the Lowell Observatory on Anderson Mesa, AZ. The NPOI comprises two sub- arrays, the Big Optical Array (BOA) and the USNO Astrometric Interferometer (AI), which share delay lines, the optics laboratory, the control system, and parts of the feed optics. We describe the design of and progress on the BOA, the imaging component of the NPOI. The AI is described elsewhere (Hutter, these proceedings). As of the date of this symposium, most of the civil engineering is complete, including the control and laboratory buildings and the concrete piers for the initial array. Three AI siderostats and associated feed pipes, three delay lines, the initial three-way beam combiner, and much of the control system are in place. First fringes are anticipated in April. By the end of 1994, four AI and two BOA siderostats, as well as three more delay lines, will be installed, making imaging with all six siderostats possible. The complete BOA will consist of six 50 cm siderostats and 30 siderostat stations in a Y with 251 m arms, with baseline lengths from 4 m to 437 m. Nearly redundant baseline lengths will allow fringe tracking on long baselines on which the visibilities are too low for detection in real time. A six-way beam combiner (Mozurkewich, these proceedings) will allow simultaneous measurements of 15 visibilities and nine of 10 independent closure phases. The output beams will feed 32-channel spectrometers covering the range from 450 to 900 nm. We anticipate tracking fringes on stars brighter than 10^m, imaging surfaces of stars brighter than 4^m, measuring stellar diameters to 0.18 milliarcsec (mas), and measuring binary orbits with major axes as small as 0.4 mas.

The USNO Astrometric Interferometer (USNOAI; the dedicated subarray of the Navy Prototype Optical Interferometer at Lowell Observatory, Flagstaff, AZ) is presently under construction and expected to begin limited operations in the spring of 1993. The main goal of the USNOAI observations is to provide a northern hemisphere catalog of approximately a thousand stars with positions known to a few mas. In order to meet this requirement, a baseline laser metrology system must be employed to measure the 3D motions of the baselines with an accuracy better than approximately equals 0.1 micrometers . The metrology scheme, as presently conceived, represents the largest and most complex high-resolution laser metrology system ever attempted.

The single R₀ Michelson interferometer beam combiner designed for use with the six array element Navy Prototype Optical Interferometer is a hybrid between the all-on-one design adopted by COAST and the pairwise design we developed for initial closure phase measurements. We describe our motivations for this design and discuss the tradeoffs between various approaches.

The U.S. Naval Observatory Astrometric Interferometer (AI) is the dedicated astrometric subarray of the new Navy Prototype Optical Interferometer (NPOI) at Lowell Observatory, which is being built in collaboration with the Naval Research Laboratory. The Naval Research Laboratory is constructing the imaging subarray of the NPOI, the `Big Optical Array' (BOA). The AI will be in operation on Anderson Mesa, near Flagstaff, Arizona, in the spring of 1994. The AI was built using the experience gained from the Mark III Interferometer on Mt. Wilson, CA, which demonstrated `proof of concept' of wide-angle astrometry by a long baseline optical interferometer. The AI incorporates four siderostats that are located in a Y-shaped configuration, and features a full-array laser metrology system to monitor baseline motion. The AI shares with the BOA state-of-the-art delay lines and a real-time zero- order fringe tracking system. The AI will have a limiting magnitude of 10 and will produce star positions accurate to a few milliarcseconds (mas). With a planned operational lifetime of several decades, this instrument will be capable of maintaining the HIPPARCOS reference frame through repeated observations, yielding improved proper motions of thousands of the brighter HIPPARCOS stars.

The ASEPS-O Testbed Interferometer is a long-baseline infrared interferometer optimized for high-accuracy narrow-angle astrometry. It is being constructed by JPL for NASA as a testbed for the future Keck Interferometer to demonstrate the technology for the astrometric detection of exoplanets from the ground. Recent theoretical and experimental work has shown that extremely high accuracy narrow-angle astrometry, at the level of tens of microarcseconds in an hour of integration time, can be achieved with a long-baseline interferometer measuring closely-spaced pairs of stars. A system with performance close to these limits could conduct a comprehensive search for Jupiter- and Saturn-mass planets around stars of all spectral types, and for short-period Uranus-mass planets around nearby M and K stars. The key features of an instrument which can achieve this accuracy are long baselines to minimize atmospheric and photon-noise errors, a dual-star feed to route the light from two separate stars to two beam combiners, cophased operation using an infrared fringe detector to increase sensitivity in order to locate reference stars near a bright target, and laser metrology to monitor systematic errors. The ASEPS-O Testbed Interferometer will incorporate these features, with a nominal baseline of 100 m, 50- cm siderostats, and 40-cm telescopes at the input to the dual- star feeds. The fringe detectors will operate at 2.2 micrometers , using NICMOS-III arrays in a fast-readout mode controlling high-speed laser-monitored delay lines. Development of the interferometer is in progress, with installation at Palomar Mountain planned to begin in 1994.

The ASEPS-O Testbed Interferometer control system design presents new challenges as compared to previous generation instruments. Increased instrument complexity due to narrow-angle astrometric observing techniques, longer baselines, and increased user expectations due to computer technology advances all contribute to the size of the effort required to bring the instrument on- line. This paper discusses the design objectives for the computer systems and software for the instrument and how the architecture selected was driven by the objectives and by the resources available; one of the major design objectives was to come up with an architecture that keeps software, network, and communications issues separate from scientific and subsystem implementation issues, making the most effective use of both scientifically- oriented and computer-oriented developers. The paper then presents the architecture that has been implemented in detail.

The Cambridge Optical Aperture Synthesis Telescope, COAST, now has the capacity to measure visibility amplitudes and closure phase for stellar sources. This paper summarizes the current status of the instrument and how the data is analyzed.

The Cambridge Optical Aperture Synthesis Telescope, COAST, is a four-telescope array for high resolution imaging using measurements of complex visibilities and closure phases. This paper describes what its component parts are and why.

The CHARA array is an optical and IR imaging array of seven 1-m aperture telescopes with a Y-shaped configuration contained within a 400-m diameter circle. The facility will be capable of submilliarcsecond imaging and will be devoted to a broad program of science aimed at fundamental stellar astrophysics in the visible and the astrophysics of young stellar objects in the infrared spectral regions. The concept for the array has been carried through Phase A feasibility and Phase B preliminary design stages with funding provided by the National Science Foundation. This paper will provide a progress report on the status of the project.

The problem of designing a beam combiner that involves more than two or three beams is complex and by no means solved. No such system has been built to date. The CHARA Array will require a beam combining system that can cope with up the seven beams with a number of spectral channels. The design of a beam combiner is probably driven more by the available technology than any theoretical constraints. Many tradeoffs between ease of manufacture and required integration time are involved. A design solution for CHARA will be presented that will incorporate spatial fringe encoding and forming fringes in the image plane.

The first two telescopes of the Infrared-Optical Telescope Array (IOTA) project are now in place and yielding data at the Smithsonian Institution's F. L. Whipple Observatory on Mt. Hopkins, near Tucson, Arizona. The IOTA collectors are 45 cm in diameter, and may be moved to various stations in an L-shaped configuration with a maximum baseline of 38 m. A third collector will be added as soon as funding permits. Each light-collector assembly consists of a siderostat feeding a stationary afocal Cassegrain telescope that produces a 10-X reduced parallel beam, which is in turn directed vertically downward by a piezo-driven active mirror that stabilizes the ultimate image position. The reduced beams enter an evacuated envelope and proceed to the corner of the array, where they are turned back along one arm for path compensation. The delay line, in one beam, consists of two parts: one dihedral reflector positioned in a slew-and-clamp mode to give the major part of the desired delay; and a second dihedral mounted on an air-bearing carriage to provide the variable delay that is needed. After delay, the beams exit from the vacuum and are directed by dichroic mirrors into the infrared beam-combination and detection system. The visible light passes on to another area, to the image-tracker detectors and the visible-light combination and detection system. The beams are combined in pupil-plane mode on beam splitters. The combined IR beams are conveyed to two cooled single-element InSb detectors. The combined visible-light beams are focussed by lenslet arrays onto multimode optical fibers that lead to the slit of a specially-designed prism spectrometer. For the visible mode, the delay line is run at several wavelengths on one side of the zero- path point, so that several cycles of interference occur across the spectrum. First results were obtained with the IR system, giving visibilities for several K and M stars, using 2.2 micrometers radiation on a N-S baseline of 21.2 m. From these measurements we obtained preliminary estimates of effective stellar diameters in the K band.

The interferometric mode of the ESO Very Large Telescope (VLT) permits coherent combination of stellar light beams collected by four telescopes with 8-m diameter and by several auxiliary telescopes of the 2-m class. While the position of the 8-m telescopes is fixed, auxiliary telescopes can be moved on rails, and can operate from 30 distributed on the top of the Observatory site for efficient UV coverage. Coherent beam combination can be achieved with the 8-m telescopes alone, with the auxiliary telescopes alone, or with any combination, up to eight telescopes in total. A distinct feature of the interferometric mode is the high sensitivity due to the 8-m pupil of the main telescopes which will be compensated by adaptive optics in the near-IR spectral regime. The VLT Interferometer (VLTI) part of the VLT Programme is conceived as an evolutionary program where a significant fraction of the interferometer's functionality is funded, and more capability may be added later while experience is gained and further funding becomes available. Major subsystems of the present baseline VLTI include: three auxiliary telescopes, three delay lines which permit combining the light from up to four telescopes, and a laboratory which contains an imaging beam combiner telescope and enough space to accommodate a number of experimental setups. This paper presents a general overview of the recent evolution of the project and its future development.

The stringent and specific requirements associated with visible aperture synthesis projects call for a sound engineering effort in the design and development phase to assess the instrumental performance. An important area of effort concerns the influence of the natural or man-made environmental factors on the global performance of the interferometer. This paper discusses the major environmental factors affecting the Very Large Telescope Interferometer (VLTI) and presents the results of a number of studies aimed at evaluating the effects of such environmental factors.

In the framework of the development of the Very Large Telescope Interferometer (VLTI) by the European Southern Observatory (ESO) we have proposed the LAMP (Large Amplitude Modulated Path). The sensor is that part of the fringe-tracker which only provides a set of error signals reflecting Optical Path Differences (OPD) at a common focus, between collected fields, the so-called piston effect. LAMP relies on both temporal encoding of OPD on each pair of apertures and on an appropriate spatial encoding which allows simultaneous operation on several apertures. Besides, it has been designed so as to provide at the same time both coherencing error signals and cophasing error signals (these latter when allowed by the source's brightness). Such ability is based on the large amplitude used in the modulation. In this paper, after a short presentation of both initial working conditions and specific capabilities that have driven the design, we describe the principle of LAMP. We also describe a generic setup and we report performances in terms of OPD resolution (minimum detectable piston), response-time and limiting magnitude. Finally advantages and drawbacks of the design are discussed.

The operation of stellar interferometers suffers from turbulence- induced random fluctuations of optical pathlength difference between collected fields. Active compensation needs an error signal which is provided by a fringe sensor. A phase A study for a fringe sensor for the ESO Very Large Telescope Interferometer (VLTI) has been conducted at OCA, leading to a proposition for the completion of a prototype. In this article, the goals and the principle of the sensor are recalled (see Gay and Rabbia, preceding paper, in this symposium). Its optimal working wavelength is discussed. Results of a numerical simulation of the sensor operation are reported, comprising sensitivity estimates. The proposed setup is then described in the details, emphasis being put on monomode optical fiber related items. Finally, current plans for the testing and the future use of the prototype are outlined.

This paper describes the present status of the Grand Interferometre a 2 Telescopes (GI2T). We review the general features of this instrument and present the scientific programs pursued by our group. Attention is given here to procedures of instrumental and visibility calibration, including the response of both the detector and spectrometer. We discuss the method of data analysis and the attainable accuracy of astrophysical parameters. The current limitations of the GI2T and development of our new optical table are presented.

We present the Active Stabilization in Stellar Interferometry (ASSI) beam combining optical table which was installed on the 2- telescope interferometer (I2T) of the Observatoire de la Cote d'Azur in 1993. To achieve very high angular resolution, the 26- centimeter telescopes can be positioned along a 140-meter North- South baseline. The limiting magnitude of the instrument depends dramatically on its ability to stabilize the fringe pattern despite the atmospheric disturbances. The function of the ASSI table is to perform this task. Three adaptive mirrors are used. The first two are fine pointing mirrors which correct the fluctuations of the angle of arrival of the two wavefronts. The other corrects the optical path difference fluctuations between the two telescopes. These corrections, e.g. tip-tilt and piston phase, are required to obtain high precision visibility measurements. We present our first observing results obtained on bright stars that have allowed the evaluation of the ASSI table performance in image tracking.

The Sydney University Stellar Interferometer (SUSI) is a long baseline optical amplitude interferometer. In its initial configuration it is a two aperture, single r_o instrument with wavefront-tilt corrections and dynamic optical path length compensation. It has been designed to measure the angular dimensions of stars of essentially all spectral types as well as the angular separations of binary stars. SUSI is located alongside the Australia Telescope at the Paul Wild Observatory, near Narrabri in northern New South Wales, Australia. It has a North-South array of input stations giving baselines covering the range from 5 m to 640 m. The baselines are being progressively commissioned, starting with the shorter ones, in parallel with an observing program aimed at fine-tuning the performance of the instrument. Progress and results from the commissioning program and the current status of the instrument are described.

The Navy Prototype Optical Interferometer (NPOI) at the Lowell Observatory near Flagstaff, Arizona is a tri-baseline stellar interferometer with specific design characteristics for astrometry and imaging. All major construction has been completed. Installation of scientific instrumentation began in July 1993 with first light expected during the Spring 1994. Here we present a description of the location, physical plan, and construction of the interferometer.

Recently, a number of technical improvements in the Infrared Spatial Interferometer (ISI) helped to increase the signal-to- noise ratio in fringe power by about a factor of two over the previously reported factor of 10. The improvements comprise higher quantum efficiency and larger bandwidth HgCdTe heterodyne detectors, better IF signal processing components, new lock-in amplifiers, a fringe calibration system and an enhanced autoguider. A comprehensive effort to characterize and improve the short-term and long-term stability of the infrared detection system led to a large improvement of the calibration of visibility data. The ISI has been used on baselines of 10- and 32-m length during the last observing season. Fringes were obtained on 8 sources on the 32-m baseline so far. This paper described recent system upgrades and a new filter bank for spectroscopy on molecular lines, as well as some more studies of atmospheric fluctuations. Results of our astrophysics and astrometry programs are reported in the following three papers.

We discuss visibility data and its analysis for 15 late-type stars observed with the U. C. Berkeley Infrared Spatial Interferometer (ISI). The ISI is a two-element heterodyne interferometer operating in the 9- to 12-micrometers wavelength region and is located at Mt. Wilson. Visibility curves were calculated using a radiative transfer model and compared with the visibility data from the ISI. A (chi) ² fitting procedure has been used to estimate the inner radii of the dust shells, optical depths at 11 micrometers , and the temperature at the inner radii. For stars in which the dust is completely resolved estimates of the stellar diameter and temperature can also be made. We present preliminary visibility data for the stars NML Cyg and IRC +10420 obtained recently using a 10-m baseline. In addition, we present preliminary recent data and analysis for the six stars, alpha Ori, omicron Cet, IRC +10216, R Leo, VY CMA, and R Aqr using a 32-m baseline. The visibility data for alpha Ori covers a sufficient range of spatial frequencies to make a determination of its diameter. Within a 95% confidence interval we obtain a diameter of 0".053+/- 0".003. For the other stars the recent 32-m baseline data are compared with previous models and further constrains some of their parameters.

We report VLBI observations at 7-mm wavelengths of the SiO maser emission in the circumstellar envelope of the late-type, supergiant variable star VX Sagitarius. Synthesis images, the first ever made of the SiO emission from VX Sgr, show that the maser emission at a stellar phase of about 0.4, arises in a ringlike distribution with a radius of about 1.3 R_*. The distribution of emission around the limb is asymmetric, as is the spectrum of emission with respect to the stellar systemic velocity. The strongest emission arises from a redshifted center of activity that lies to the south of the star and may indicate an asymmetry in the stellar atmosphere or in the mass loss. The maser emission within 4 km s^-1 of the stellar velocity is distributed along the limb and does not show evidence of systematic velocity gradients greater than a few km s^-1. However, there may be evidence of acceleration of material away from the star, where outflow velocities increase from about 10 km s^-1 at 1.3 R_* to at least 20 km s^-1 in the OH and H₂O maser shells at radii of about 30 R_*. The arrangement of the maser emission supports a model having dense velocity coherent structures with characteristic sizes of approximately 0.5 AU in the extended stellar atmosphere. However, a substantial fraction of the maser flux density has been resolved by these observations, as with earlier VLBI observations of at least three times more coarse angular resolution. We speculate that the circumstellar shell has velocity coherent cells on spatial scales of approximately 0.5 to perhaps 100 AU, which give rise to SiO maser emission. The SiO maser emission lies well within the 4.6 R_* inner radius of the surrounding dust shell, as measured with the U. C. Berkeley infrared interferometer. The high angular resolution infrared and millimeter observation were made at similar phases during the same cycle, and thus provide strong evidence that the SiO maser emission arises well away from the dust-rich region of the circumstellar envelope.

Infrared astrometry at the 10-milliarcsecond (mas) level is applicable to experiments in stellar evolution astronomy, solar system dynamics, relativistic gravitation, and deep space laser tracking. We are pursing astrometry with the U. C. Berkeley Infrared Spatial Interferometer (ISI) on Mt. Wilson to demonstrate a 10-mas capability for tracking stellar and solar system objects. Astrometric data from the ISI, taken and analyzed over the last 5 years, have shown that instrumental and atmospheric effects limit current demonstrations. The ISI data show that point-to-point interferometric phase fluctuations due to tropospheric and quantum noise, for optimal integration times of 0.2 seconds, are approaching the 0.1-cycle level needed to reliably connect the phase. Modeling the ISI data suggests that atmospheric fluctuations on Mt. Wilson, during the best seeing, are dominated by a low-lying component, within the first 25 meters above the ISI, which, in the future, may be minimized with in situ calibration. A calculation of atmosphere-limited astrometric accuracy shows that the ISI will soon be able to achieve 10-mas astrometry, on a 13-m baseline in a single observing session, employing current ground-based laser distance interferometer calibrations to minimize atmospheric effects.

LATOR is a space-based experiment to accurately measure the gravitational deflectional deflection of light. The experiment uses two laser bearing spacecraft at the opposite side of the Sun and a very long baseline heterodyne interferometer to measure the angle at an accuracy of 0.2 uas. Combining this measurement with laser ranging from Earth to both spacecraft, gravitational deflection can be made with an accuracy 5000 times better than previously done and will allow measurements of the second order and frame dragging effects.

The SAO Submillimeter Wavelength Array is under construction and is expected to be ready for observations in late 1997. It will consist of six telescopes with baselines from 9 to 470 m and resolutions as fine as 0.'1. Eight receivers will cover all bands from 180 to 900 GHz in linear polarization; two orthogonal polarizations will be available at 350 GHz. Multiyear site testing on Mauna Kea shows that the precipitable water vapor is less than 1 mm for 17% of the total time, and the rms phase fluctuations on a baseline of 100 m are less than 55micrometers for 25% of the total time. The configuration of the array will be optimized to provide nearly uniform coverage in the uv plane within a circular boundary for an instantaneous observation of a source at the zenith. To achieve this distribution, the antennas will be placed along the sides of Reuleaux triangles. The basic electronic architecture follows conventional practice in radio interferometry and involves heterodyne frequency conversion. The correlator-spectrometer will have 92000 spectral channels (6144 lags per baseline) and will be able to provide 1.6-MHz resolution for a total bandwidth of 4 GHz. There were several significant design changes in the past year: a sixth mirror was added to the optical path in each antenna; the antenna configuration was changed to four tangential rings rather than four concentric rings; the correlator rings; the correlator clock rate was increased from 40 to 52 MHz and the IF conversion scheme was changed from a single-sideband conversion to two conventional mixer conversions, in order to improve the image rejection.

The Smithsonian Astrophysical Observatory (SAO) is constructing an array of antennas that is expected to be placed on Mauna Kea and operated at submillimeter wavelengths. To facilitate calibration work in single dish mode, each antenna is equipped with a subreflector that chops about its center of mass. In addition to chopping, the subreflector's position is controlled by three actuators permitting the mirror position to be corrected for the effects of temperature and pointing elevation. This paper discuses the specifications, design, fabrication, and performance testing of the chopping subreflector system.

The Mark III Interferometer on Mt. Wilson, a long-baseline optical interferometer, was in daily operation for more that seven years. During that time it achieved milliarcsecond angular resolution for binary star astronomy, with submilliarcsecond accuracy. For the first time many spectroscopic binaries have been resolved, including binaries in which the companion cannot be detected with spectroscopy. The high angular resolution means that the traditional gap between visual and spectroscopic binaries has been decreased by more than an order of magnitude. In order to confirm the performance of the Mark III Interferometer, this paper uses the results of astronomical observations, and compares the Mark III Interferometer with other high-resolution techniques, including astrometry, lunar occultation, photometry, speckle, and spectroscopy. Comparisons for a variety of binary stars among these techniques indicate that long baseline optical interferometry proves a reliable, fully automatic, daily accessible astronomical capability for achieving high resolution, high accuracy, high dynamic range, and high photometric measurement precision for the study of binary stars.

Standard FFT-based phase screen generation methods do not accurately model low-frequency turbulence characteristics. This paper introduces a new phase screen generation technique which uses low frequency subharmonic information to correct the problem. We compare our technique to two other subharmonic methods. The structure functions for this new method match very closely the structure functions of Kolmogorov turbulence theory.

This paper deals with the analysis of optical measurement data on atmospheric turbulence characteristics. Some conclusions are drawn on the variability of the most large-scale component of the turbulent inhomogeneities, spectrum for the atmosphere as a whole and for the boundary atmospheric layer as well. We will analyze several years of experimental results.

Bester et al. report measurements of atmospheric fluctuations made with the Infrared Spatial Interferometer, which indicated behavior not in accord with the standard Kolmogorov model with only a single constant wind velocity. Our numerical simulations use relatively complex models of the atmosphere to investigate both Kolmogorov and non-Kolmogorov models. We find that the measurements of Bester et al. for light passing through the upper atmosphere are within the limits of behavior for Kolmogorov models, but often only if the outer scale of turbulent fluctuations is between 15 to 100 meters. The possibility that the measured behavior might be non-Kolmogorov is not excluded. We also examine measurements made along short paths in the surface boundary layer, where some measurements of Bester et al. showed variations in the atmospheric fluctuations with seeing conditions which appeared to be non-Kolmogorov. These variations can be perhaps be explained by standard models, but require that seeing improve with increasing wind speed in the surface layer. We discuss some other measurements which lend some support to that idea. However, we cannot exclude non-Kolmogorov behavior. We find that meteorological data is needed concurrent with astronomical observations, to help constrain the models. The size of the outer scale, the wind velocity profile and the turbulence spectrum are important to the ultimate capabilities of interferometers and other systems with adaptive optics.

We have used interferometric methods to track the atmospheric phase fluctuations at two astronomical sites: the Observatorio del Roque de los Muchachos in the Canary Islands, the site of the 4.2-m William Herschel Telescope (WHT), and the Lord's Bridge observatory in Cambridge, the location of the Cambridge Optical Aperture Synthesis Telescope (COAST). At both sites the atmospheric perturbations are well characterized by Kolmogorov turbulence. At the WHT we find no obvious evidence for saturation of the turbulence over the range of spatial scales investigated, although there are rapid variations in the seeing on timescales of minutes.

In order to design adaptive optics systems and to study the effect of atmospheric turbulence on astronomical images it is helpful to know the temporal power spectra of the various Zernike polynomial terms. These coefficients correspond directly to standard aberration terminology like tilt, astigmatism, spherical aberration and so on. An analytical method based on work by Roddier et al. for predicting these power spectra using the current average wind speed, Frieds parameter r_(Omicron ) and the aperture size is presented along with some experimental results testing their predictive power. The data are also shown to confirm the important results of Fried as well as Noll.

Traditional differential astrometric techniques are limited in precision by the atmosphere in a way that does not show much improvement with increased telescope aperture. However, greatly improved astrometric precision may be obtainable by exploiting the strong aperture dependence of the spatial correlation between simultaneously recorded specklegrams within the speckle isoplanatic angle. The cross-correlation of two speckle iages of a binary star pair may yield higher astrometric precision in the measurement of the binary separation than centroid differences. The degree of this improvement, however, depends strongly upon the effective thickness of the turbulence in the atmosphere. A 5- minute observation using a large-format, rapid-readout CCD at a 2.3-m telescope has demonstrated 1-milliarcsec precision in the determination of the separation of a 7.3 arcsec binary star pair when processed with speckle techniques.

Designs for the large binocular telescope (LBT) beam combiner optics, collimators, and reimaging cameras for the visible and infrared over a range from 0.5 to 13 microns are presented and discussed. To achieve good images at the combined focus of the LBT, an adaptive correction will be made independently for each of the two telescopes prior to beam combination. For interferometry the image planes of each telescope are first collimated. The two beams then pass to two flats. The two flats will be shifted to remove the remaining piston differences between the two telescope beams, using an error signal derived from a fringe tracker further down the optical path. The flats output a pair of parallel collimated beams with the collimator exit pupils and separation scaled to the original LBT entrance pupil geometry. Because the pixel matching, diffraction limits and isoplanatic fields are different in the different wavelength bands, a series of cameras have been designed covering the different bands from the visible through 12 microns. Both refractive and reflective optical designs are considered.

We present interferograms and reconstructed images obtained with the MPE imaging beam combiner simulator COSI.The purpose of COSI is to simulate the imaging beam combiner at the coherent focus of the ESO VLTI in multi-speckle mode of under conditions of partial or full correction of the single telescope wave front by adaptive optics. The data discussed here were taken in multi-speckle and single speckle mode. COSI consists of a 1-m telescope and a near- IR continuum light source to simulate the radiation from astronomical objects. Two flat mirrors allow us to use one half of the telescope as a transmitter and the other half as a receiver. In the receiving focus we have installed the MPE speckle camera SHARP, which uses a HgCdTe 256² NICMOS 3 array. A pupil mask over the aperture allows us to simulate various telescope configurations like the one with a beam compression factor of 100 as it will be used for the ESO VLT interferometer. COSI is used to explore NIR array detector properties and their suitability for interferometric measurements and to generate data to develop and test image reconstruction algorithms. First interferograms of single and multiple objects were taken early 1993. Employing various deconvolution and Fourier-inversion methods, a diffraction limited image of the pin-hole sources can be successfully recovered which experimentally demonstrates the feasibility of interferometric imaging with a large monolithic beam combiner. Thus, we have demonstrated that COSI is an excellent test bed to investigate methods of image recovery and to investigate how the methods are influenced by effects like atmospheric turbulence, expected optical imperfections and detector characteristics. Early this year (1994) we installed hot-air seeing simulators for individual subapertures. This allows us now to take interferograms in multi- speckle mode.

We present the layout and construction of main parts of the imaging beam combiner simulator experiment COSI, which has been set up at the Max Planck Institut fuer extraterrestrische Physik. The purpose of COSI is to simulate the imaging beam combiner at the coherent focus of the ESO VLTI. Main objectives of the experiment are to explore near-infrared array detector properties and to generate data for testing (developing) image reconstruction algorithms. COSI consists of a 1-m telescope and a near-infrared continuum light source to simulate the radiation from astronomical objects. Close to the focal plane, the beam is split into two foci. The beam is relayed between the two parts of the telescope by a roof mirror. Between the telescope and the retroreflector in the quasi-parallel beam, a pupil mask allows us to define a number of beams, thus simulating various telescope configurations with a beam compression factor of 100. The light from the pin-hole source is transmitted through the telescope system to the near-infrared camera placed in the receiving focus. The wavefront in the beams can be influenced by seeing simulators. Presently we use hot air devices, but for the final configuration we aim at deformable mirrors to have detailed control over the wavefront. Experimental results on image recovery achieved up to now are reported elsewhere in this conference.

Experiments of the optical interferometer were made with two coude telescopes of 25-cm aperture which are set up by 2.6 m apart in the north-south direction. Interferometry by artificial lights were carried out. The fluctuations of fringes of He-Ne laser correlate to vibrations of the equipments. Fringes of alpha Boo were observed. However, the fringes were detected only a little because of the insufficient accuracy of delay compensation and the phase fluctuation due to the atmospheric turbulence.

With the advent of speckle interferometry and similar techniques, it has been possible to obtain orbital data, and, thereby, masses for a number of binary stars which were higherto unresolvable. For the full astrophysical exploitation of this benefit, it then becomes essential to obtain photometric and spectrophotometric information on these systems. Photometry can often be secured by extensions of the basic speckle resolution technique, and several solutions have been proposed for the problem of obtaining uncontaminated spectra of the normally blended stellar components such as objective prism speckle spectroscopy and wideband projection speckle spectroscopy, and passive interspectroscopy. We present here further practical development of the methods suggested in Paper I and the results of some observational tests of the technique.

We present the results of some laboratory experiments of the use of electro-optical (EO) devices to control the optical path length (OPL) of an interferometric array. One of the most important problems in interferometric beam combination is the control of the path length; this is coupled with the need for partial wavefront compensation in order to increase the sensitivity of the interferometer. Traditional approaches to such problems are often very expensive and sometimes impractical. For this reason we started an effort, both theoretically and experimentally, in order to investigate if less costly and more effective techniques can be applied. In our experiments we used single-cell LCDs in order to eliminate piston terms in a two- aperture interferometer. We used phase diversity techniques for extracting the phase information. Although the experimental results are still partial we believe that there is enough evidence that such devices can be used for the OPL control and partial wavefront compensation. Further testing is needed in order to assess the real capabilities of commercially available LCDs and the need, if any, of customization.

Imaging array interferometry in the optical regime offers the promise of observing faint sources of small angular extent. A ground based system, however, must ameliorate the effects of atmospheric turbulence to maintain the very high resolution imaging capabilities of the instrument. Observing these interesting sources will require large diameter collectors with partial wavefront compensation. We highlight the mathematical treatment expressing the ensemble average image spectrum for the image plane and pupil plane interferometers. The theory treats aperture co-phasing and wavefront compensation by the idealized removal of an arbitrary number of Zernike modes.

We present computer simulations of variable baseline 2D imaging optical interferometers operating at visible and infrared wavelengths. Sparse apertures of fewer than 10 receivers, baselines up to 400 m and aperture size from 1 to 2 m are considered. SNR limitations pose significant problems with dilute apertures observing faint sources; we explore various ways to address this problem. We simulate pupil plane visibility measurements under perfect conditions and in the presence of wavefront aberrations due to atmospheric distortion. A variety of ideal sources are studied, including stellar photospheres with features and artificial satellites. Earth rotation aperture synthesis over extended periods improves spatial frequency coverage of astronomical sources, while observation at multiple wavelengths improves coverage for geosynchronous satellites. We introduce a hybrid technique for performing a fast analytic pixel transform, similar to the fast discrete Fourier transform, which allows complex sources to be represented in pixel form but which admits the full floating point accuracy of analytic calculation. We study image deconvolution techniques to enhance the final image. An algorithm is presented for improvement of images formed from pupil plane interferometric data. Values are added to the frequency domain with the dominant constraint being an image taken with a small filled aperture instrument. A deconvolution technique using the entropy function is developed to enhance reconstruction of the truth image. Examples of image improvement obtained by the algorithm are presented.

In this paper we describe an instrument to obtain diffraction limited images on a large telescope using aperture masking. Detailed images of giant stars have been routinely obtained on the 4.2-m William Herschel Telescope (WHT) in La Palma with this technique. When using this method one of the major causes of data loss is the difficulty in detecting the secondary mirror support spiders crossing, and thus obscuring part of the aperture mask. To overcome this problem a system for continuously monitoring the front face of the aperture mask has been devised. Exposure times have been significantly reduced by placing the image in a corner of the CCD chip, the size of which is determined by the seeing, and by compressing the image to a single row of pixels. In order to perform this technique the fringes must be accurately aligned with the columns of the CCD chip. To eliminate the extra losses in throughput due to an optical image rotator the CCD camera, rather than the image, is rotated using a motorized mount. The aperture mask can be rotated independently to various position angles with respect to the object so that full UV coverage can be obtained.

We examine the effects of low-order adaptive optics on the performance of closure-phase (nonredundant mask) imaging. In particular we investigate a system that uses a small number of plane tip-tilt correcting mirrors. These apply local adaptive corrections to the beams transmitted by a pupil plane mask, but do not correct for piston wave-front errors between the beams. We have identified the optimum sub-aperture size and science channel exposure time for a range of seeing conditions and light levels using numerical simulations. Our calculations predict improvements in the SNR of monochromatic power spectrum and bispectrum measurements by factors in the range 2-15 over their uncorrected values. For optical NRM imaging with large telescopes, this combination of sparse localized wavefront correction, rather than modal correction over the whole telescope aperture, is much more efficient in terms of maximizing the SNR improvement per adaptive degree-of-freedom. Such a zonal AO system represents a cost-effective strategy for extending the utility of closure phase imaging using a very small number of active optical components. Suitable astrophysical sources amenable to such an approach include evolved supergiant stars, long period variables, and other bright compact objects.

Problems are considered referring to development of a large adaptive telescope consisting of individual modules with common image and synthesized aperture. The telescope capabilities are estimated by Fresnel number as product of aperture size by field dimensions divided by wavelength. It is shown that configurations based on `power' modules are most promising ones to provide superlarge values of Fresnel number rather than conventional configurations with afocal modules. Some problems are solved including image matching by using of the images curved surfaces having the common center and simultaneous slope of the modules towards the telescope center. Problem of phasing across the field is solved by introducing spherical aberration into pupils of the modules. A power grating interferometer configuration is suggested, too, usable as a sensor for the synthesized wavefront deformations just during observation of an extent source. As an example, a multimodule telescope configuration is proposed with Fresnel number of up to 5(DOT)10⁴.

A holographic system for phasing of multimodule synthesized aperture optical telescope is proposed. A phasing test system model is theoretically analyzed for an implementation being composed of three flat transmissive diffraction gratings with identically oriented and equidistant lines of diffraction hologram structures. The proposed holographic system irradiation by an infinitely far nonmonochromatic source is shown to result in the telescope focal plane as a noncoherent addition of monochromatic fringe pattern of the same phase and orientation produced by the beams having passed through the telescope modules. A criterion for phasing accuracy estimation is proposed being ratio of fringe pattern phase increment measured in any point of the telescope focal plane to the irradiation wave length change which has caused this increment. Mathematical expression is obtained describing the dependence of the achieved phasing accuracy on tolerance values for the modules geometrical parameters.

The paper deals with the problem of choosing proper wavelength for a wavefront sensor involving an internal laser source and a hologram on the surface of the primary mirror of a telescope. Some results of dimensional calculations for the optical arrangement for recording the hologram structure are presented. As a result a conclusion is drawn on the possibility of reducing the size of an optical stand for making this structure by increasing the operation wavelength of the control channel of a telescope.

A summary of the scientific requirements and the optical design requirements for the Large Binocular Telescope (LBT, former Columbus Project) beam combining optics is presented. Our goal for LBT is to produce a phased focal plane combining the beams from two 8.408 m diameter primaries on a common mount (14.417 m center-center). This provides an interferometric baseline of 22.825 m (edge-edge). This combined focal plane should be phased, in-focus and unvignetted over a field covering most or all of the isoplanatic patch in the best seeing conditions. These field requirements should be met for wavelengths from 0.4 to 20 micrometers . This allows the observer to obtain the maximum amount of diffraction-limited imaging information around each reference source. The individual telescope images should be close to diffraction-limited over the isoplanatic patch in order to facilitate adaptive correction of the interferometric field. A minimum number of warm reflections and minimum obstruction of the beams should be used in order to reduce the emissivity. The two- shooter telescope will be uniquely powerful in the thermal infrared (based on the combination of interferometric baseline and single element diffraction limit). We will be discussing several related classes of beam combiners to evaluate design tradeoffs for LBT. An earlier design considered direct combination of the beams from two F/33 secondary mirrors with long back focal distance. We have also considered reimaging the F/15 Gregorian focal planes with a number of optical designs.

We present applications of a recently developed iterative blind deconvolution algorithm to both simulated and real data. The applications demonstrate the algorithm's performance for a wide range of astronomical imaging. We demonstrate the effectiveness of using multiple observations of the same object convolved with different point spread functions. We also show the extension of the algorithm to phase retrieval when the object Fourier amplitude is available.

The Mark III optical interferometer has been in routine operation on Mt. Wilson, near Los Angeles CA, since 1988. Because it employs active fringe-tracking, seeing measurements are a natural byproduct of any astronomical observation. An automated seeing analysis program has allowed us to extract measurements of the coherence time, t_(omicron ), and the high-frequency spectral index of the power spectrum of fringe motion for all clear nights in the observing seasons 1989-1991. This provides a unique database of the seeing; very few other measurements of the temporal properties of astronomical seeing exist, and none with such a representative sampling of nights. We examine various statistical properties of the seeing and infer quantities of interest to the design of future interferometers.

A new focal instrumentation for the Grand Interferometre a 2 Telescopes (GI2T) called REGAIN (REcombinaison pour GrAnd INterferometre) is under study at the Observatoire de la Cote d'Azur (OCA) and the Laboratoire d'Astronomie Spatiale (LAS) in Marseille, France. The objectives of the REGAIN project are multiple. Priority number 1 is a more efficient astrophysical exploitation of the GI2T. Next is the possibility for observing simultaneously at visible and near-infrared wavelengths. Finally REGAIN should ensure the test of the OVLA prototype telescope added to the present GI2T. Therefore, the resulting GI3T could be used for phase-closure imaging with 1.5-m apertures. At the same time reservations will be made for implementing adaptive optics units for each telescope whilst the VLT interferometer fringe- sensor currently studied at the OCA, should be tested on the GI3T.

We describe a concept for an interferometric space mission dedicated to global (wide-angle) astrometry. The GAIA satellite contains two small (baseline APEQ 3 m) optical interferometers of the Fizeau type, mechanically set at a large and fixed angle to each other. Each interferometer has a field of view of about one degree. Continuous rotation of the whole satellite provides angular connections between the stars passing through the two fields of view. Positions, absolute parallaxes and annual proper motions can be determined with accuracies on the 20 micro-arcsec level. The observing programme may consist of all objects to a limiting magnitude around V = 15-16, including 50 million stars. The GAIA concept, which has been proposed for a Cornerstone Mission within the European Space Agency's long-term science programme, is based on the same general principles as the very successful ESA Hipparcos mission, but takes advantage of the much higher resolution and efficiency permitted by interferometry and modern detector techniques.

Differential speckle interferometry is based on the cross analysis of series of speckle patterns produced in different wavelengths. The study of the position differences between these speckles provides angular information on objects much smaller than the diffraction limit. In order to make the measurements of the photocenter displacement, we have built an instrument which behaves like a spectrograph in one direction and a speckle interferometer in the perpendicular direction. A mirror anamorphoser permits us to meet the different sampling requirements. The dispersed speckle pattern is recorded by a photon counting camera. The measurements of the photocenter displacements are very sensitive to differences of aberration between spectral channels and temporal variations of the detector's distorsion. Our instrument provides images with a quality equal to the diffraction limit plus residual aberrations of the order of one hundreth of the wavelength used. The distorsion of the optics is much smaller than the size of the temporal variations of the detector's distorsion. In order to correct this variable distorsion, spatial and spectral modulations are made in a fully automated instrument.

We discuss the performance of a fiber-coupled image-plane interferometer in which one transports individual image-plane wavefronts point by point by means of bundles of single-mode fibers to a beam-combining station where such wavefronts are cross-correlated. We compute the SNR of image reconstruction under low-light conditions where the principal source of noise is the shot noise of photodetection.

In this paper we propose the usage of optical correlation techniques as an analog real-time correlation tracker that can be applicable to solar observations or any other complex image stabilization situation. We present the theory of the VanderLugt and joint extimate filters for optical correlation. Some straightforward scheme of implementation of such devices will be given. Furthermore we will discuss the main advantages for such correlators mainly compared with more traditional digital-based methods.