This paper aims to give a broad view of the progress achieved in ground-based interferometry over the past ten years and to assess quantitatively the factors determining the types of object that can be observed with high resolution over the next ten.

The GI2T interferometer has been recently equipped with a new beam combiner called REGAIN, including a dedicated visible spectrograph for enhanced spectral capabilities. The control system, the fringe tracking processor, the photon counting detectors as well as the data reduction pipeline have been refurbished or changed. After a long phase of qualification, the whole system is now ready for scientific operations. We will describe the main new features of the system.

We present a summary of the status of the Cambridge Optical Aperture Synthesis Telescope, and review developments at the array through the period 2000-2002. Summaries of the astronomical and technical programmes completed, together with an outline of those that are currently in progress are presented. Since our last report two years ago in 2000, there have been significant changes in the context for astronomical interferometry in the UK. We review these developments, and describe our plans for the near and intermediate term at COAST, and with colleagues in Europe at the VLTI and in the USA at the Magdalena Ridge Observatory in New Mexico.

SUSI -- the Sydney University Stellar Interferometer -- currently operates with a maximum baseline of 160 m. A major upgrade which will see the sensitivity of the instrument increased by a factor of 50 is nearly complete, and first fringes have been obtained with the new red beam-combining system.

The U.C. Berkeley Infrared Spatial Interferometer is a two telescope stellar interferometer operating in the 9-12 micron atmospheric window, utilizing heterodyne detection with CO₂ laser local oscillators. Science with the ISI has been focused on the measurements of the spatial distribution of dust and molecules around mass-losing late type stars, and more recently precision measurements of stellar diameters in the mid-infrared avoiding molecular lines. During the past few years, a National Science Foundation sponsored program of expansion from two to three telescopes has been underway. This expansion will allow the ISI to make visibility observations on three simultaneous baselines and a measure a closure phase. The third telescope was completed last year and shipped to Mt. Wilson, and more recently a Central Control Facility and Master Laser Oscillator Facility were also completed and recently shipped to Mt. Wilson. In this paper we report progress on this program and highlight some of the most recent astrophysical results.

New beam combination techniques, using two and three telescopes, have been the focus of activity at IOTA during the past two years since our last update. In particular, we have added a third telescope, made closure-phase measurements, demonstrated two- and three-beam combination with integrated optics combiners, demonstrated two-beam combination with an asymmetric coupler, and made simultaneous JHK visibility measurements with an image-plane combiner.

This paper describes the current status of the Navy Prototype Optical Interferometer with emphasis on imaging and on changes since the last review in this forum.

We discuss recent work from the Palomar Testbed Interferometer (PTI), including science results and system improvements. In the past two years PTI has been used to observe a wide range of scientifically interesting sources, including binaries, Cepheids and Miras. In addition PTI has been used to observe departures from spherical symmetry in several stars. Recent system improvements incude a new low read-noise camera based on a HAWAII infrared array, routine opteration in two baselines, and operation in the J band. Future developments include an upgrade to three-aperture combination and closure phase measurements, and double-Fourier interferometry.

The CHARA Array is a six element optical and near infrared interferometer built by Georgia State University on Mount Wilson in California. It is currently operating in the K and H bands and has the largest baseline (330 m) in operation of any similar instrument in the world. We expect to begin I band operations in 2002. We will present an update of the status of the instrumentation in the Array and set out our plans for the near term expansion of the system.

The Keck Interferometer combines the two 10 m Keck telescopes for high sensitivity near-infrared fringe visibility measurements, nulling interferometry at 10 μm to measure the quantity of exozodiacal emission around nearby stars, and differential-phase measurements to detect "hot Jupiters" by their direct emission. First fringes with the interferometer were obtained in March 2001 using the two Kecks with their adaptive optics systems. Subsequent engineering work has been focused toward the visibility mode in the areas of system validation, and improving sensitivity, increasing automation, and adding functionality in preparation for nulling and differential phase. Recently four shared-risk teams were selected by NASA to participate in early science observations, and initial shared-risk science observations have begun.

The Very Large Telescope (VLT) Observatory on Cerro Paranal (2635 m) in Northern Chile is approaching completion. After the four 8-m Unit Telescopes (UT) individually saw first light in the last years, two of them were combined for the first time on October 30, 2001 to form a stellar interferometer, the VLT Interferometer. The remaining two UTs will be integrated into the interferometric array later this year. In this article, we will describe the subsystems of the VLTI and the planning for the following years.

Mitaka optical and InfraRed Array second phase instrument, MIRA-I.2, has obtained the first fringes on 8th June 2002 at a baseline of 30m and the second fringes on 13th August. MIRA-I.2 consisting of two 30cm siderostats was proposed and started designing in 1994. Major constructions started in April 1999 just after the first phase instruments MIRA-I.1 was closed. The telescopes were initially put at a 6m test baseline in the MIRA-I.1 dome and obtained fringes in June 2001 and then moved to the 30m baseline in August 2001. At present, some of parts do not reach the final shape. Observations of star radii or binary orbits will be planed, while remained construction will be continued.

The Large Binocular Telescope (LBT), with dual 8.4 m optics on a common mount, is unique among the large-aperture interferometers. Deformable secondaries on the telescope capable of adaptive atmospheric correction allow beam combination after only three warm reflections. The design allows the implementation of two powerful uses of interferometry: suppression of starlight (or nulling interferometry) and wide-field imaging (or Fizeau interferometry). Nulling will allow detection of extrasolar planetary systems (from either zodiacal emission or giant planets) down to solar system-equivalent levels for nearby stars. This will dramatically increase our knowledge of the prevalence and make-up of extrasolar planetary systems. Fizeau interferometry will allow imaging of even complex structure at the resolution of a 22.8 m telescope. To implement these two powerful techniques the University of Arizona and NASA are collaborating to build the Large Binocular Telescope Interferometer (LBTI) a cryogenic instrument capable of sensitive interferometric observations in the infrared.

The Magdalena Ridge Observatory (MRO) project is presently funded to design and build a facility including an optical/infrared imaging interferometer composed of up to 10 1.5 meter class telescopes and a single conventional 2.4 meter class telescope. The interferometer array will be arranged in a “Y” configuration and the use of movable telescopes will allow its reconfiguration from a very compact array with baselines up to tens of meters to a true long baseline configuration with baselines up to 400 meters. We plan to introduce adaptive optics systems on the array telescopes.

We introduce for the first time an approach to interferometer design which makes explicit the global nature of the optimization of the array. We discuss the idea of a "fitness landscape" for interferometric arrays, where a fitness function is defined in terms of the success of the array, and the coordinates of the space are design parameters such as numbers and sizes of telescopes, baseline length and wavelength coverage. We investigate the extent to which such ideas can illuminate our understanding of strategies for array design and our ways of evaluating the success of a design.

When multiple, adaptively-corrected telescopes are coherently combined at a common focal plane, interference appears as spatial structure in the point spread function (PSF). True images of deep fields can be reconstructed at resolution λ/b from multiple images recorded to cover the u-v plane with baselines up to b. Assuming no losses or detector noise, sensitivity improves with the number of elements, n, and of combined images, N. The formalism of PSF fitting is used here to determine signal/noise ratio for faint point sources seen against sky background noise. For two-element and dilute, non-redundant arrays the PSF fit is dominated not by any sharp peak, but by fringes or weak speckles, and the signal/noise ratio is proportional to square root N(2n-1). For close packed or highly redundant arrays the sensitivity increases as nsquare root N, and is the same as for a single dish with the same total collecting area and integration time. It follows that the signal/noise ratio for a two-element array is thus square root 3/2 = 86% of this maximum limit. We show that the requirement for PSF sampling with negligible detector read noise can be met with available optical and infrared array detectors, for baselines up to 10 times the element diameter. Low-loss combination could be realized in practice by an interferometer with two large moving elements, with baselines set up sequentially to sample the full u-v plane. In the 20/20 concept, two 21 m telescopes move continuously around a circular track during an integration, to keep the baseline oriented perpendicular to the source. Coheret combination is made at a station held midway between, to obtain aperture synthesis of images covering the full, one arc-minute field corrected by multiconjugate adaptive optics. With 16 images taken over 8 hours to fully cover baselines up to 100 m, the resolution in the K band is 4 mas, and the 10σ limiting magnitude for point sources will be 27.7.

The 8-m class telescopes are now in full operation, while 100-m baseline interferometers (VLTI, KeckI) are starting routine operation too. A working group from the French high angular resolution community tried to identify what could be our post-VLT/VLTI instruments after 2010. Possible future instruments, ground or space-based, can be split into three main categories: Extremely large filled aperture telescopes, diluted interferometric arrays for direct imaging, and diluted interferometric arrays for aperture synthesis imaging. These concepts are compared in terms of observing capabilities and performances (spatial resolution, field of view, imaging capability, sensitivity, photometric dynamical range, etc.), technological issues (adaptive optics, phasing, instrument mount, etc.) and R&D priorities.

The Jean-Marie Mariotti Center is a network of 11 French Institutes, Laboratories or Observatories, appointed by CNRS in 2000. It coordinates the efforts of the member institutes to offer all the potential users of interferometric facilities the best operational environment, providing software, academic formation and stimulating the prospective on new interferometric developments. At present, besides academic formation, the major effort is focused on the development of the software to prepare the observations, to reduce the data and to interpret the results in terms of models or reconstructed images. In this contribution, we describe the achievements and the future plans of the Mariotti Center.

The start of NEVEC was initiated by the opportunity in the Netherlands to reinstate instrumental efforts in astronomy through a funding program for 'Top Research Schools,’ which brought about the creation of NOVA. The fact that considerable experience exists in Radio Astronomical imaging through interferometry (the Westerbork Synthesis Radio Telescope started in 1970), and the relatively small size at the time of ESO's VLTI Team made it opportune to aim for a win-win situation through collaboration. So presently an MOU between ESO and NOVA is in force, which stipulates that 10 out of the 18 man-years funded by NOVA for NEVEC until 2005 [new personnel, in university setting (Leiden) but on project money] shall be used on tasks that are mutually agreed between NOVA and ESO. The tasks presently are found in the domain of observing modes, calibration and modeling, as well as contributing to the commissioning of new instruments and thinking about future instruments. Another task, outside these 10 FTE, has been the data handling and analysis software for MIDI, and again contributing to its commissioning. Delivery of the first operational version in Heidelberg has just taken place (summer 2002) contributing to the successful Preliminary Acceptance in Europe for MIDI on September 10, 2002. The actual state of 'products and deliveries' and the future outlook are reviewed.

To coordinate the on-going work of the various interferometric groups at German institutions the German Center for Infrared and Optical Interferometry, called FrInGe, was created in September 2001. The center will coordinate and support the German activities in obtaining, reducing and interpreting astronomical interferometric data from optical to mid-infrared wavelengths. The center will keep a publication archive, an interferometric data base, and will carry out tutorials for training of the next generation of astronomers in optical and infrared interferometry. In addition, FrInGe has established cooperations with other interferometric centers in Europe.

We present observations of several T Tauri stars using long baseline infrared interferometry from the Palomar Testbed Interferometer. The target sources, T Tau N, SU Aur, RY Tau and DR Tau, are all known to be surrounded by dusty circumstellar disks. The observations directly trace the inner regions (<1 AU) of the disk and can be used to constrain the physical properties of this material. For three of the sources observed, the size scale of the infrared emission is tenths of AU, which is considerably larger than predicted by flat disk models. We discuss the implications of these results for models of circumstellar material, in particular the recent theoretical work suggesting the presence of an extended vertical wall at the inner edge of the disk.

We present observations of the symbiotic star CH Cyg with a new JHK-band beam combiner mounted to the IOTA interferometer. The new beam combiner consists of an anamorphic cylindrical lens system and a grism, and allows the simultaneous recording of spectrally dispersed J-, H- and K-band Michelson interferograms. The observations of CH Cyg were conducted on 5, 6, 8 and 11 June 2001 using baselines of 17m to 25m. From the interferograms of CH Cyg, J-, H-, and K-band visibility functions can be determined. Uniform-disk fits to the visibilities give, e.g., stellar diameters of (7.8 ± 0.6) mas and (8.7 ± 0.8) mas in H and K, respectively. Angular stellar filter radii and Rosseland radii are derived from the measured visibilities by fitting theoretical center-to-limb intensity variations (CLVs) of Mira star models. The available HIPPARCOS parallax of CH Cyg allows us to determine linear radii. For example, on the basis of the K-band visibility, Rosseland radii in the range of 214 to 243 solar radii can be derived utilizing CLVs of different fundamental mode Mira models as fit functions. These radii agree well within the error bars with the corresponding theoretical model Rosseland radii of 230 to 282 solar radii. Models of first overtone pulsators are not in good agreement with the observations. The wavelength dependence of the stellar diameter can be well studied by using visibility ratios V(λ₁)/V(λ₂) since ratios of visibilities of different spectral channels can be measured with higher precision than absolute visibilities. We found that the 2.03 μm uniform disk diameter of CH Cyg is approximately 1.1 times larger than the 2.15 μm and 2.26 μm uniform-disk diameter.

We present bispectrum speckle interferometric observations of the deeply embedded protostellar outflow source S140 IRS1. Using the SAO 6 m telescope, we obtained a K-band image with diffraction-limited resolution of 76 mas, which is the highest resolution image of a young outflow source ever obtained in the infrared. Our image shows the circumstellar environment of S140 IRS1 in unprecedented detail and suggests that the central source is marginally resolved with a FWHM diameter of approximately 20 mas (approx 20 AU). The dominant feature is a bright extended and very clumpy structure pointing away from the central source in exactly the same direction as the blue-shifted CO outflow lobe. We interprete this feature as the clumpy inner surface of a partially evacuated cavity in the circumstellar envelope around IRS1, which has been excavated by the strong outflow from IRS1. In addition, we find several arc-like structures north-east of IRS 1, extended diffuse emission south of IRS 1, and four new point sources. The diffuse and fragmentary structures close to IRS 1 appear to trace circumstellar material swept up by energetic outflows. In combination with molecular line emission maps from the literature, our image provides direct confirmation that two distinct bipolar outflow systems continue to be driven from IRS 1 on scales between 3" and 100". Our speckle observations provide important complementary information for future long-baseline interferometric observations, for example with the LBT.

We present near-infrared (JHK) bispectrum speckle-interferometry monitoring of IRC+10216 obtained with the SAO 6m telescope. The present speckle observations covering baselines up to 6m provide important complementary informations for future long-baseline interferometry. To disentangle the apparent motions of the various IRC+10216 components and to reveal the location of the central star, future high-resolution observations are of utmost value for the interpretation of this astrophysical key object. The J-, H-, and K-band resolutions of our speckle observations are 50 mas, 56 mas, and 73 mas, resp. The K-band observations cover 8 different epochs from 1995 to 2001 and show the dynamical evolution of the dust shell which consists of several compact components within a 200 milli-arcsecond radius. Our recent two-dimensional radiative transfer modelling has shown that the central star is probably not located at the brightest dust-shell component A but at the position of the northern component B. The bright and compact component A is the southern lobe of a bipolar structure. The changes of the dust-shell structure can be related to corresponding changes of the optical depth caused, for instance, by mass-loss variations. The present observations are consistent with the predictions of hydrodynamical models that enhanced dust formation takes place on a timescale of several pulsational cycles.

Individually resolved packets produced by scans from the CHARA Interferometer Array for binary stars can be analyzed in terms of the astrometry of the binary without using visibilities. We considered various methods for finding the locations of the packets, including autocorrelation and Shift-and-Add, but our best results were obtained from a method of direct packet fitting. This method was put to use in analyzing two data sets each for the stars 12 Persei and Beta Arietis respectively. These data were taken between Nov 6 and 15, 2001 with the CHARA Array 330 m E1-S1 baseline. Some 460 to 830 scans were taken in both directions with the auxiliary PZT, and seeing conditions were fair to poor for these runs (r₀ ≈ 7 cm). This procedure yielded a projected separation for each data set, with an intrinsic accuracy of 0.15 - 0.3 mas. This represents an order of magnitude improvement over speckle interferometry techniques. The orbits were refined by a maximum likelihood technique. In the case of 12 Per the semimajor axis obtained was α = 53.53 mas, compared with the previous orbit of 53.38 mas, a small increase of 0.27%, which implies a mass increase of 0.8%, an insignificant change for this well-established orbit. For Beta Arietis, we find that α = 35.62 versus the previous orbit's value of 36.00 mas. This is a 1.0% decrease, resulting in a mass decrease of 3.0% for this system.

We report the first long-baseline interferometric observations of R CrB. The observations were carried out at the Infrared Optical Telescope Array (IOTA), using our new JHK beam combiner which enables us to record fringes simultaneously in the J-, H-, and K-bands. The circumstellar envelope of R CrB is resolved at a baseline of 21 m, and the K-band visibility is derived to be 0.61 ± 0.03 along a position angle of about 170 degrees. The visibility obtained with IOTA, as well as speckle visibilities with baselines up to 6 m and the spectral energy distribution (SED), are fitted with 2-component models consisting of the central star and an optically thin dust shell. The K-band visibilities predicted by the models are about 10% smaller than the visibility obtained with IOTA. However, given the simplifications adopted in our models and the complex nature of the object, this can be regarded as rough agreement. As a hypothesis to explain the small discrepancy, we propose that there might be a group of newly formed dust clouds, which might appear as a third visibility component.

At Mitaka campus of National Astronomical Observatory of Japan, we are now constructing the MIRA-I.2 Interferometer, the second stage interferometer of MIRA (Mitaka optical and InfraRed Array Project). The MIRA-I.2 system is an interferometer consisting of two 300 mm siderostats which are placed on a 30 m baseline. Fringe detection is made at 800 nm with APD. Before setting up the regular system of 30 m baseline, we made the test system of 6-m baseline. This system was used for testing and solving the problems of the remote control of fast tip-tilt mirrors, the remote control of siderostats, and the remote adjustment of optics. After the fringe detection of stars with this system, in June 2001, we complete the regular system of MIRA-I.2 of 30 m baseline. In August, 2001, we started the setup of the regular system of 30 m baseline of the MIRA-I.2 interferometer by making the modifications and improvements of each element. From the end of October 2001, the test observations of some bright stars have been made. The fringes of α Lyr were detected with the regular system, in June 2002. Object stars, such as single star, binary star, Cepheid, Mira variable to be observed are investigated. Main subjects of this interferometer system are the observations of star radius and binary star.

During the 2001 observing season, the CHARA Array was in regular operation for a combined program of science, technical development, test, and commissioning. Interferometric science operations were carried out on baselines up to 330 meters -- the maximum available in the six-telescope array. This poster gives sample results obtained with the approximately north-south telescope pair designated S1-E1. At operating wavelengths in the K band, the 330 m baseline is well suited to diameter determinations for angular diameters in the range 0.6 - 1.2 milliarcseconds. This is a good range for study of a wide range of hot stars. In this poster, angular diameters for a set of A,B and F stars are compared to results derived from other sources. These confirm CHARA performance in the range 3-10% in visibility. The normal stars follow a normal spectral type - surface brightness relation, and a classical Be star deviates from the norm by an amount consistent with its apparent colors.

We show two samples of raw data from the Navy Prototype Optical Interferometer (NPOI), the first with a single baseline present, the second with three baselines present. We focus on the behavior of the group delay amplitude and the position of the group delay peak, and compare those quantities in the data samples to simulated data. We find that the simulations resemble the actual data when the flux level and spectral regions that were observed are realistic, and conclude that the NPOI fringe engine seems to be working properly.

We report here the first visibility and closure-phase measurements done with the IONIC instrument at the IOTA interferometer. The IONIC instrument is presented and preliminary analysis of the results discussed. Future improvements of IONIC are envisioned.

We have developed an algorithm to determine precise fringe phases in the presence of atmospheric turbulence. We use phase bootstrapping to improve parameter estimates of weak fringes observed on long baselines through coherent integration. With data from the Navy Prototype Optical Interferometer (NPOI) on γ Sagittae, we demonstrate the importance of this method for the study of limb-darkening of stars.

In March 2001, the commissioning instrument of the VLTI, VINCI, succeeded in obtaining its first fringes by linking two 40 cm aperture siderostats on a 16 m baseline. During the first year of operation, thousands of interferometric observations on different baselines were carried out, with the technical goal of characterizing this complex system. We report in this paper these first measurements and estimate the main parameters of the atmospheric and internal turbulence along the complete light path. We first illustrate the degradation of the visibility accuracy caused by the differential piston and evaluate the contribution of the internal optical path fluctuations with respect to the atmospheric ones. The stability of the VLTI complex is demonstrated, which enabled us to record easily fringes with Unit Telescopes (UTs) on baselines as long as 102.5 m (November 2001). In the last part, infrared measurements of the atmospheric differential piston are reported. They were obtained with the siderostats on two different baselines ranging from 16m to 66m. Estimations of the coherence time at Cerro Paranal are derived from these commissioning data and compared to the values predicted by the Astronomic Site Monitor (ASM). Finally, constraints on the outer scale length are discussed.

Single-mode fibers have been used in the near-infrared to dramatically reduce calibration error for long-baseline interferometry. We have begun an effort to apply the advantages of spatial filtering at visible wavelengths for precision measurements of pulsating Cepheids using the IOTA interferometer. Rather than employing photometric taps to calibrate fluctuating coupling efficiency, we are using an "asymmetric" coupler which allows this calibration to be done without losing photons. The Single-Mode Asymmetric Recombination Technique (SMART) experiment has finished lab-testing, and has been installed at IOTA for full commissioning in Summer 2002. We report the results of lab characterization and first sky tests, as well as first fringes on a star using a visible-wavelength single-mode coupler. With both lab and sky experience using unpolarized light, we have found that circular silica fibers are quite practical for precision interferometric measurements. We conclude that circular fibers (as opposed to polarization maintaining fibers) have an undeserved poor reputation and that birefringence effects pose no significant difficulty.

AMBER, Astronomical Multi BEam combineR, is the near-infrared focal instrument dedicated to the VLTI. It is designed to combine three of the VLTI Telescopes and to work simultaneously in the J, H and K spectral bands (1.1 to 2.4 μm). The instrumental concept and its opto-mechanic specifications were defined in order to reach the ambitious scientific requirements to satisfy the core astrophysical programs. The project passed the Final Design Review in May 2001, phase which marks the acceptation of the instrument final design and the beginning of the construction and tests. After this phase, optics and mechanical systems have been receptioned since February 2002, for the laboratory tests and alignments. The cooled spectrograph and its cryostat is assembled at the Osservatorio di Arcetri in Firenze, Italy and the cooled detector at the Max-Planck-Institut fur Radioastronomie in Bonn, Deutschland. The warm optics, including spatial filter in K and artificial sources injection system, have been pre-aligned at the Observatoire de la Cote d'Azur in Nice, France to validate most of the alignment procedure, the required element accuracies and the associated degrees of freedom. The whole instrument is then currently fully assembled and optimized at the Laboratoire d'Astrophysique de l'Observatoire de Grenoble, France. Its sensitivity and final performance will be evaluated in order to reach the Preliminary Acceptance in Europe, scheduled beginning 2003. This paper gives the results of the warm optics laboratory commissioning.

This paper will provide details of the outstanding performance of the VLTi Delay Lines in Blind Tracking and Fringe Tracking modes. This OPD control performance was demonstrated during the first fringe event in October last year where two UTs (ANTU and MELIPAL) were combined by the interferometer. Already in March 2001 the fringes were observed by the VLT interferometer equipped, at this time, with two Siderostats. The stellar light was collected and trasported via the Delay Lines to the VINCI instrument, also used as fringe detector. The excellent results were achieved after several missions on the mountain, managed by ESO and DUTCH SPACE personnel, to install and to verify the tracking performance of the first three Delay Lines over the full length of the tracks and to align the science beams over the Delay Lines reflectors. The first performance tests were done in open loop (without a fringe sensor feedback to the Delay Line) by imposing a global trajectory on the active Delay Line. During the commissioning period with the Siderostats, this procedure was a success and provided already encouraging results. In a later stage of the Delay Line commissioning, the Fringe Acquisition and Tracking Mode enabling to keep closed loop performance for longer periods each night replaced the Blind Tracking Mode. Performance monitoring during various occasions after installation revealed that OPD errors are not degrading in time; extensive preventive maintenance activities are not required to keep the performance levels within specifications, proving the robustness of the controller of the Delay Line.

As the result of an analysis pursued from the very beginning, today the VLT Interferometer is the only interferometer allowing to have a 2 arcsec interferometric field of view (f.o.v) available at the instruments entrance. This accessible interferometric field is the direct result of a careful pupil transfer from the individual telescopes to the central laboratory, unique feature of the VLTI. For this goal it has been necessary to develop a new optical device, the Variable Curvature Mirror (VCM.), using large deformation theory of elasticity, and advanced techniques in optical fabrication. The possibility with the VLTI to use various baselines, from 8 to 200 m with UTs or ATs, leads to severe conditions on the VCM curvature range. A given delay-line, and its associated VCM, should be able to transfer a pupil to the interferometric laboratory from a very far or relatively close position of an ATs. Considering the f.o.v required in the VLTI (2 arcsec), the delay-lines strokes or the OPD to compensate for, and the various locations of the UTs and ATs stations, the curvature of the VCM has to be continuously variable within a range from 84 mm^-1 to 2800 mm^-1. The location of the VCM in the delay-line system, on the piezo-translator used for small OPD compensation, led to minimize its dimensions and to realize a small active mirror with a 16mm diameter. With this small optical aperture, the VCM range of curvature corresponds to a f ratio from f/∞ to f/2.625. The two first VCM complete systems (mirror, mechanics and control command software) have been achieved in 2001/2002 and will be installed in the VLTI delay-lines during fall 2002. Their final performances (optical quality, pupil transfer accuracy, etc.) are reviewed.

We present a summary of the global system analysis that led to the current definition of the AMBER instrument. AMBER is a near infrared multi-beam combiner for the Very Large Telescope Interferometer. This analysis goes through the following issues: atmospheric systematics including atmospheric turbulence and dispersion, analysis of single mode optical fibers, photometry calibration, spectral dispersion, background noise, data reduction and calibration steps.

MIDI is the Mid-Infrared interferometer for ESO's VLTI (Very Large Telescope Interferometer), which has been developed by a German-Dutch-French consortium [MPIA Heidelberg Germany, NOVA/ASTRON Dwingeloo Netherlands, Observatoire de Meudon France]. The initial aim of MIDI is to combine the beams from 2 telescopes in the 10 micron N-band with a spatial resolution of up to 10 milli-arcseconds and a maximum spectral resolution of 230. Modulation of the optical path difference can be done using piezo-driven mirrors at room temperature, but beam combination and detection of the interferometric signal has to be done at cryogenic temperatures due to the 'thermal' wavelength domain. The MIDI cold bench is therefore mounted inside a cryostat, cooled by means of a closed cycle cooler to about 40K for the cold optics and 8K for the detector. The design of the cold optics has been kept as simple as possible, creating challenges such as preserving alignment from 295K to 40K and accessibility. This poster describes the realization of the cold optics, the alignment and test strategies and laboratory results.

In its current configuration, the VLT Interferometer (VLTI) combines the light collected by two telescopes and directs it towards the commissioning instrument called VINCI. In an interferometer, the optical path ranging from a telescope to the point where beams are combined is referred as an arm of the interferometer. This arm contains a large number of optics that have to be aligned at installation time and kept aligned during the period of use of the interferometer. The method used to perform the initial alignment is reported in a separate article. This paper is focussed on the methods used to assess the stability of the image alignment of each interferometer arm. Collected data sets are presented and interpreted.

A remote operations center for Georgia State University's Center for High Angular Resolution Astronomy (CHARA) Array is in the final stages of implementation on the university campus in Atlanta, GA. Several technological considerations were incorporated into the overall design including a secure network infrastructure with an acceptable end-to-end latency, a control room replete with appropriate computing and projection systems, an efficient client-server model, and a data archival system. Although independent of the local weather, remote operations have practical considerations, such as routine preparations requiring on-site personnel and the observation of astronomical targets with celestial coordinates appropriate to the Local Sidereal Time (LST) and U-V plane coverage of the array.

In this paper we describe the telescope optics, manufacturing tolerances and the geometric alignment procedure of the CHARA telescopes. We also report on our efforts to test and refine the alignment of the telescopes by implementing the curvature sensing method. The results of the first experiments on telescope W1 show that we can get consistent results with this method. We also found a slight distortion caused by the lateral support of the primary mirror.

The CHARA Array consists of six 1-meter telescopes. The telescopes are at fixed positions laid out in a Y-shaped pattern, where the longest available baseline is 330 meters. The resolving power of this interferometric array operating at visible and short infrared wavelengths is better than one milli-arcsecond. The current infrared beam combination system is capable of combining the light from any two of the six telescopes in the array. With the existing infrared beam combination and detection system, we routinely observe in K and H band, where our magnitude limit is 6.

The CHARA Array at Mt. Wilson uses a PICNIC array camera for fringe detection, connected to a realtime fringe tracking computer running RTLinux. This paper describes the PC- and RTLinux-based camera controller and software that is used to allow high-speed, deterministic, low-latency readout of frames from the camera, as well as a camera simulator that mimics the behavior of the camera. This camera controller is built from commercial off-the-shelf (COTS) PC hardware and uses software running on the free RTLinux operating system, resulting in a very inexpensive camera controller system. The hardware costs for the system, including the PC (although excluding the costs of analog signal interfaces and power supplies), are less than $2000. The controller is capable of reading out arbitrary subimages from the camera, can quickly switch between different readout patterns, and is capable of controlling either CCD cameras or infrared array cameras. Detailed camera timing can be supplied by and/or tuned by the end user, as desired. In addition, a camera simulator unit has been developed. This camera simulator allows the development of camera interface hardware without the risk of damage to the expensive camera. The camera controller described connects to the Niro camera supplied to CHARA by Mark Shure, and the camera simulator mimics the behavior of this camera.

The design and scientific objectives of a near infrared channeled spectrometer planned at the IOTA interferometer are discussed. The spectrometer has the flexibility to reconfigure easily for conventional broadband operations in addition to multi-channel mode. This instrument makes use of the existing PICNIC camera at the IOTA in order to be cost efficient. The spectrometer has been designed specifically for studying Mira stars. However, it will find its application in other areas of astrophysical interests such as studies of circumstellar disks around young stars and binary stars.

We describe the array metrology system of the Navy Prototype Optical Interferometer, a long-baseline interferometer whose purpose is to determine precise positions of bright stars and image the surfaces and circumstellar environments of stars and stellar systems. For the astrometric array of the NPOI to achieve its design goal of wide-angle astrometric precision of 1-3 milliarcseconds on 20 m baselines, an extensive laser metrology system is employed to measure the three-dimensional motions of those baselines with respect to an Earth-fixed reference system to an accuracy of approximately 100 nm. The array metrology system is described, along with its associated data analysis software, test results, and the application of those results to the analysis of astrometric data. It is shown that at the current stage of its development, the array metrology system is capable of monitoring the variations in the geometry of the astrometric array with micron-level precision.

A key thrust of NASA's Origins program is the development of astronomical interferometers. Pursuing this goal in a cost-effective and expedient manner from the ground has led NASA to develop the Keck Interferometer, which saw first fringes between the twin 10m Keck telescopes in March of 2001. In order to enhance the imaging potential of this facility, and to add astrometric capabilities for the detection of giant planets about nearby stars, four 1.8 m 'outrigger' telescopes may be added to the interferometer. Robust performance of the multi-aperture instrument will require precise alignment of the large number of optical elements found in the six optical beamtrains spread about the observatory site. The requirement for timely and reliable alignments dictated the development of an automatic alignment system for the Keck Interferometer. The autoaligner consists of swing-arm actuators that insert light-emitting diodes on the optical axis at the location of each optical element, which are viewed by a simple fixed-focus CCD camera at the end of the beamtrain. Sub-pixel centroiding is performed upon the slightly out-of-focus target spots using images provided by a frame grabber, providing steering information to the two-axis actuated optical elements. Resulting mirror-to-mirror alignments are good to within 2 arcseconds, and trimming the alignment of a full beamtrain is designed to take place between observations, within a telescope repointing time. The interactions of the autoaligner with the interferometer delay lines and coude trains are discussed in detail. The overall design of the interferometer's autoaligner system is presented, examining the design philosophy, system sequencing, optical element actuation, and subsystem co-alignment, within the context of satisfying performance requirements and cost constraints.

When completed the VLTI project will be composed by four 8.2 m Unit Telescopes (UT) and four 1.8 m Auxiliay Telescopes (AT) with their respective Coude trains and relay optics, two test siderostats, 6 (up to 8) Delay lines and 8 Beam compressors with their corresponding feeding mirrors. There will be more than 200 optical components, mirrors and lenses, with diameters ranging from 5 mm to 8200 mm. Their surface shapes range from flat to off-axis ellipsoid, including also spherical, on and off-axis hyperbolae and parabolas as well as cylindrical surfaces. Depending on the interferometer configuration, the different possible optical path lengths are of the order of 100 to 300 meters. We describe briefly the principles chosen as well as the types of criteria and method used for the alignment. The method can certainly be applied to other optical systems. The explanations given are understandable to the non-optician, this text is not intended to be an alignment procedure.

The I. Physikalische Institut of the University of Cologne is participating in an international collaboration with the Max-Planck-Institut fur Astronomie in Heidelberg and the Osservatorio Astrofisico di Arcetri for the development of LINC/NIRVANA, the Near-Infrared/Visible Interferometric Camera for the Large Binocular Telescope (LBT). LINC/NIRVANA will be one of the two interferometric camera systems of the LBT and will operate at wavelengths from 0.6 μm to 2.4 μm, with the long wavelength regime between 1.0 μm and 2.5 μm being covered by LINC (LBT INterferometric Camera} and the shorter wavelengths part from 0.6 μm to 1.0 μm being processed by NIRVANA (Near-InfraRed/Visible Adaptive iNterferometer for Astronomy}. The main contributions of the Cologne institute to this camera will be the 77K dewar and the Fringe and Flexure Tracker (FFT) for the near-infrared part on the system. Detecting and correcting the fast pistonic aberrations of the atmosphere and the slow flexure of the instrument in a closed-loop operation, the presence and proper function of the FFT is mandatory for a time-stable image quality at highest interferometic resolutions. In order to get the best possible image correction for LINC, the FFT will be located inside the camera dewar at an interferometric focus close the one of the near-infrared science detector. Using simple optical elements it will continuously monitor the time-variable phase difference and pupil locations of the incoming wavefronts from the two arms of the twin-telescope. In this article we give a short overview of the camera concept of LINC and present the current status of the design and development of the FFT going on at our institute at the University of Cologne.

We propose a millisecond of arc optical/infrared array for stellar territory, MIRA-ST, with nine 4m-aperture off-axis telescopes, whose maximum baseline length is about 600 m. MIRA-ST will have the photon collecting area equivalent to that of a single-dish telescope of 12 m diameter, and the imaging capability better than 1 millisecond of arc resolution at 2.2 micrometers with a high dynamic range of reconstructed images. Combining the light beams from each pupil telescope efficiently is one of the most difficult tasks. We compare the relative merits among a so-called pair-wise beam combining, an all-on-one beam combining, and a tree-structured beam combining. As for transferring the beams from individual telescopes to a beam combining facility with the loss of photons as small as possible, an optical fiber system is a most interesting substitute for the current mirror-and-vacuum-pipe combination. Specifically, the nature of spatial filtering of optical fibers has been under study in the light of deepening the limiting magnitude attainable without introducing an adaptive optics to each telescope. With MIRA-ST we will be able to zoom in the stellar territory to unveil the detailed structures and lifecycles of stars of various kinds, and to examine the universality and/or diversity along the coarse of their evolutionary paths. The specific targets of most interesting for us are, among others, T Tauri stars, AGB and post-AGB stars, Cepheids, brown dwarfs, white dwarfs, stellar atmosphere/envelope of low temperature stars, accretion disks, and fundamental structures of main sequence stars.

The 'OHANA (Optical Hawaiian Array for Nanoradian Astronomy, means "family" in Hawaiian) aims at making a large and sensitive optical/IR array with the Mauna Kea 3 to 10 meter telescopes. Telescopes will be linked with single-mode fibers to carry the coherence of the beams from the output of the telescopes adaptive optics systems to the beam combination units. The project has been divided into three phases. The first phase is dedicated to the injection of light into single-mode fibers and to the building of the injection module. The third phase is the realization of the complete array and its use by a wide community of astronomers. In the second phase, a prototype 'OHANA will be built and the "shortest" baselines will be explored. The baselines will be located in the South-East and West parts of the observatory. An extra baseline will possibly link the two groups of telescopes if infrastructure comply with it. This phase II 'OHANA will already be the longest and most sensitive optical/IR interferometer built. Scientific targets will span young stellar objects, extragalactic sources and other types of astronomical topics which require both high angular resolution and sensitivity. This paper reviews the main characteristics of the phase II interferometer.

Once the proof of concept of the OHANA Array has been demonstrated, the Phase II capabilities can be put into regular science operation, and the OHANA facility can be upgraded to extend interferometric operation to include all of the telescopes of the OHANA Consortium member observatories. This will constitute the Phase III of OHANA. The technical developments required will be relatively straight-forward. Longer fiber sets will be procured (fiber losses are not a limiting factor at the OHANA scale). An enhanced delay line capability will be needed in order to exploit longer baselines with good sky coverage and ample super-synthesis (several compact, multi-pass long optical delay concepts are under investigation). The scheduling and operation modes of an instrument such as OHANA present interesting opportunities and complications. We envision a place for both collaborative consortium science, based on mutual allocation of facility access, and PI-driven access, based on telescope access exchange between consortium members. The most potentially successful mode of operation would imply a community driven model, open to proposals from the different time allocation comittees. This poster looks at possible methods of allocation and operation, inspired by the UKIRT infrared survey (UKIDSS), the European VLBI, and the very interesting possibility of a Mauna Kea telescope time exchange scheme. The issue of data property is of course intimately tied with the proposal/operation system, and means of data availability and distribution are discussed, along with data interpretation tools, which may be modeled on existing systems such as the ISC at Caltech or the JMMC in France. when weighed against the UV coverage, the potential science and the uniqueness of this project, all these issues are worth an in depth study. Discussions are starting as to an OHANA Operation Committee, the goal of which would be to discuss, define and eventually carry out operational modes. The goal, of course, is for the Operation Committee to handle the details of multi-telescope scheduling in a way that will be transparent to the scientist who merely seeks the observational results.

The next generation of optical interferometer arrays will require a large number of unit telescopes in the same manner as the VLA if meaningful scientific objectives are to be achieved. Studies based on the five element COAST array show that something like ten to fifteen telescopes are necessary. For such a project to be viable the unit telescopes must be designed from the outset for this task. The basic criteria are as follows: The wavefront quality and stability should be excellent, high optical throughput, autonomous automatic operation, couple efficiently into the beam transport and combination system, plus maintain acceptable unit cost. To achieve these goals a number of novel designs were considered and are described in this paper. Two of the most suitable designs and which had the least technological risk were studied in more detail by Telescope Technology Ltd. and are described in a separate paper.

The OHANA interferometric array will be implemented by linking existing Mauna Kea telescopes with optical fiber. No new facility construction on Mauna Kea is required or planned for OHANA. Fibers will be run through existing cableways. The maximum potential baselines are approximately 800 meters in length. Interferometric operation with good UV coverage will require within the instrument variable optical delay approaching 400 meters. It will be necessary to provide this length of delay within a modest amount of existing laboratory space. An obvious approach is the use of multiple passes within a short delay line space. This poster investigates possible multi-pass implementations and related issues of efficiency, cost, wavefront quality and diffraction. The required optical delay can be provided at reasonable efficiency and moderate cost. The simpler optical delay for OHANA Phase II, already under construction, is described.

The role of the tip-tilt mirror in the stellar interferometer is introduced. Tip-tilt mirror construction, work theory and mechanical design are discussed in detail. According to the facilities we already have had, we designed two sets of measurement system to measure two important parameters: tip-tilt mirror movement range and its frequency response. The result was given which shows good quality of this tip-tilt mirror we design.

Model-predicted and observed properties of the brightness distribution on M-type Mira disks are discussed. Fundamental issues of limb-darkening and diameter definition, of assigning observational data to diameter-type quantities and of interpreting such quantities in terms of model diameters are outlined. The influence of model properties upon interpretation of measured data is clarified. The dependence of the center-to-limb variation (CLV) of intensity on wavelength, on stellar parameters and on variability phase and cycle may be used for analyzing the geometrical and physical structure of the Mira atmosphere, for determining fundamental stellar parameters, and for investigating the quality of models. Desirable future observations include simultaneous observations in different spectral features at different phases and cycles, observation of the position of the shock front and observation of the time- and wavelength-dependence of deviations from spherical symmetry.

The size and variability of the photospheres of several late-type stars has been probed using 11 micron heterodyne interferometry. High resolution observations performed during the years 1999 - 2001 yielded diameter measurements accurate to about 1% for α Ori and o Cet, an supergiant and a mira variable. Narrow bandwidths (0.17 cm^-1) and high resolution spectra were used to avoid molecular lines. Observations were made at several different wavelengths, sometimes purposely overlapping an observed spectral feature. In all cases, the 11 micron sizes are larger than previously measured visible and near-infrared diameters. The discrepancies will be discussed. In addition, a variation of the diameter of Mira with phase has been observed.

Our new IOTA JHK-band beam combiner allows the simultaneous recording of spectrally dispersed J-, H- and K-band Michelson interferograms. In this paper we present our IOTA observations of the Mira star T Cep with this beam combiner (observations in June 2001; four baselines in the range of 14 m to 27 m). The beam combiner optics consists of an anamorphic cylindrical lens system and a prism. From the interferograms of T Cep we derive the visibilities and the J-, H-, and K-band uniform-disk diameters of 14.0 ± 0.6 mas, 13.7 ± 0.6 mas and 15.0 ± 0.6 mas, respectively. Angular stellar filter radii and Rosseland radii are derived from the measured visibilities by fitting theoretical center-to-limb intensity variations (CLVs) of different Mira star models. The available HIPPARCOS parallax (4.76 ± 0.75 mas) of T Cep allows us to determine linear radii. For example, from the K-band visibility we derive a Rosseland radius of 329_-50/⁺⁷⁰ solar radii if we use the CLVs of the M-models as fit functions. This radius is in good agreement with the theoretical M-model Rosseland radius of 315 solar radii. The comparison of measured stellar parameters (e.g. diameters, effective temperature, visibility shape) with theoretical parameters indicates whether any of the models is a fair representation of T Cep. The ratios of visibilities of different spectral channels can be measured with higher precision than absolute visibilities. Therefore, we use the visibility ratios V(λ₁)/V(λ₂) to investigate the wavelength dependence of the stellar diameter. We find that the 2.03 μm uniform-disk diameter of T Cep is about 1.26 times larger than the 2.26 μm uniform-disk diameter.

Ten bright Miras and eight semi-regular variable giants and supergiants have been observed with the IOTA (Infrared and Optical Telescope Array) interferometer in the L' band (from 3.4 to 4.1 microns). Observations were carried out in March and November 2000 with the FLUOR/TISIS instrument. Variations in the diameter of R Leo are for the first time observed in the L' band, and our data show that the diameter of α Ori is remarkably stable. Important deviations from a uniform disk model are demonstrated for most of the Mira stars of our sample, and are particularly obvious for χ Cyg, R Cas and ο Cet. Observations of α Her carried out in March are consistent with previous estimates of its diameter published with the very first data of TISIS. The instrument has shown its reliability and the good quality of its data. The present results emphasize the importance of such data for a better comprehension of the circumstellar environment of evolved stars.

We show examples of PHOENIX stellar atmosphere model simulations for data obtained by long-baseline optical interferometers. A single spherical, hydrostatic model atmosphere for the M-type giant star γ Sagittae is shown to be in good agreement with interferometric, spectrophotometric, and high-dispersion spectroscopic data sets. In particular we show that the interferometric triple products and wavelength-dependent uniform disk diameters measured by the Navy Prototype Optical Interferometer (NPOI) are well matched by our model. With expanding atmosphere models for the A-type supergiant α Cygni we predict that the center-to-limb intensity profile is sensitive to the mass-loss rate of its stellar wind. For one possible NPOI configuration we show the sensitivity of the squared visibility amplitudes to the model mass-loss rate at spatial frequencies beyond the first null. The importance of hot star photospheric interferometry is discussed.

The IOTA (Infrared Optical Telescope Array) has been routinely operating with two-telescopes since 1994, a mode destined to become obsolete following its recent conversion to a three-telescope array. In two-telescope mode, the IOTA has made numerous scientific and technical contributions, see e.g. our list of publications at http://cfa-www.harvard.edu/cfa/oir/IOTA/PUBLI/publications.html. We present preliminary results on three different topics using recent data from the two-telescope IOTA: (1) measurements of Mira star diameters simultaneously in three different near-infrared spectral bands, (2) measurement of the characteristic size and shape of the source of near-infared emission in the x-ray binary system CI Cam, and (3) aperture synthesis of the Carbon star V Hydrae combining data from the IOTA and from aperture masking at the Keck-I telescope.

We report on direct interferometric measurements of stellar intensity profiles obtained with the Navy Prototype Optical Interferometer (NPOI) and the Very Large Telescope Interferometer (VLTI). These measurements need to make use of weak fringes, i.e. low visibility values, on resolved stars. We describe techniques that were used to obtain these low visibility values with high precision. They include the methods of baseline bootstrapping and wavelength bootstrapping, as well as, lately, coherent integration using phase bootstrapping. In addition, we developed methods to compensate photon and detection biases. We present recent measurements on the giant star γ Sge, obtained with the NPOI, which succeeded not only in discriminating between uniform disks and limb-darkened disks, but also in constraining Kurucz model atmosphere parameters. We present first VLTI measurements of visibility values beyond the first minimum which were taken on the giant star Ψ Phe. Here, the capabilities to synthesize baselines of different lengths and to use different aperture sizes were used for the first time with the VLTI. We close with an outlook on the future potential on studies of stellar surface structure with the six-way beam combination at the NPOI and with the completed VLTI. This includes for instance direct measurements of the limb-darkened profiles of a large number of different types of stars, and of starspots which may for instance be caused by magnetic fields or large-scale photospheric convection.

A dedicated program of measuring angular sizes of Mira variables using the Palomar Testbed Interferometer has resulted in more than 13,000 measurements of over 60 stars. With visibility data in five channels across the K band, atmospheric extension and sources of opacity are evident in the dataset. Using multiple-epoch narrow-band data, phase lags between wavelength-dependent angular size cycles indicate spatial extent of molcular atmospheres. A spectral angular diameter classification system is developed, based on the recurring shapes of the spectral traces seen in the dataset, which correlates to spectral type. A period-radius relationship is explored, and a refinement on the estimation of angular sizes is made.

The VLT interferometer has been operating since the time of first fringes in March 2001 with a pair of 40 cm diameter siderostats at baselines of 16 and 66 m and a pair of 8 m diameter telescopes (UT1 and UT3) with a baseline of 102 m using the test camera VINCI operating in the K band. A fair fraction of its commissioning time has been devoted to observing a number of objects of scientific interest around the southern sky bright enough to allow high precision visibilities to be obtained on a routine basis. A large number of stellar sources with correlated magnitudes brighter than K approximately 6 and K approximately 3 with the 8 m and 40 cm telescopes respectively have been observed over this time period with limited, u,v plane coverage. In this paper, the most interesting results on sources never observed before at these spatial resolutions and on known sources for which the VLTI data allow the establishment of tighter constraints on theoretical models will be reviewed.

Direct diameter measurements of Cepheid variables are used to calibrate the Barnes-Evans Cepheid surface brightness relation. More than 50 separate Cepheid diameter measurements from four different optical interferometers are used to calculate surface brightnesses as a function of magnitude and color. For two Cepheids, η Aquilae and ζ Geminorum, high precision diameter measurements as a function of pulsation phase are available from the Palomar Testbed Interferometer (PTI). Relations using only these diameters are found for each individual Cepheid in order to search for differences between Cepheids of different pulsation period. In all cases the best-fit relations are simple linear relations between surface brightness and color with the constraint that for a spectral type A0 star (where all colors equal zero) all relations must yield the same surface brightness (i.e., there must be a common zero-point). The derived relations found using interferometric Cepheid diameters are consistent with functions in the literature found using interferometric observations of non-variable giant and supergiant stars. In addition, while the separate relations for η Aquilae and ζ Geminorum are marginally consistent within the errors they do differ in the direction predicted for Cepheids of differing pulsation period. Using these new surface brightness relations the distance is calculated to the nearby Cepheid δ Cephei for which a new distance has been found using trigonometric parallax with the Hubble Space Telescope. These distances are well within the errors of the distance derived from trigonometric parallax.

The importance of guided optics in stellar interferometry is discussed. Specificities, advantages and drawbacks are presented.

We present details of the design and construction of a new pinhole Spatial Filtering system at the Cambridge Optical Aperture Synthesis Telescope and show the results from the first astronomical observations with the system. We discuss the advantages of using such a system to reduce uncertainties in visibility observations with an optical/IR interferometer. We compare simultaneous observations with and without the system and show the potential to improve both the quality and quantity of scientific output of an interferometer. We also show that filtered observations are much more robust to changes in atmospheric seeing than unfiltered observations, improving the accuracy of calibrated visibility measurements while reducing the need for repeated calibration. Given the simplicity and high throughput of pinhole spatial filters we suggest that they could be a valuable addition to many current and future arrays.

Wavefront cleaning by single-mode fibers has proved to be efficient in optical-infrared interferometry to improve calibration quality. For instance, the FLUOR instrument has demonstrated the capability of fluoride glass single-mode fibers in this respect in the K and L bands. New interferometric instruments developed for the mid-infrared require the same capability for the 8-12 μm range. We have initiated a program to develop single-mode fibers in the prospect of the VLTI mid-infrared instrument MIDI and of the ESA/DARWIN and NASA/TPF missions that require excellent wavefront quality. In order to characterize the performances of chalcogenide fibers we are developing, we have set up an experiment to measure the far-field pattern radiated at 10 μm. In this paper, we report the first and promising results obtained with this new component.

The fibered beam combiner FLUOR, which has provided high accuracy visibility measurements on the IOTA interferometer, is being moved to the CHARA array which provides five 1m telescopes on baselines ranging from 35 to 330m. The combination CHARA/FLUOR makes it possible for the first time to achieve sub-milliarcsecond resolution in the K band, with a dynamic range of 100 or more. We explore the scientific potential of CHARA/FLUOR, most notably in the domains of high contrast binaries and the characterization of Cepheid pulsations, and present some of the anticipated developements.

Developments of fiber linked optical interferometer are reported. This interferometer is a part of MIRA-I.2 interferometer (Mitaka InfraRed and optical Array). MIRA-I.2 is an optical interferometer with a 30 meters long baseline. It consists of two 30cm siderostats, tip-tilt mirrors, vacuum pipes delay lines and detectors. We plan to use two 60 meters long polarization-maintaining fibers for arms of the interferometer, instead of vacuum pipes. The developments include dispersion and polarization compensation of fiber and fiber injection module. In laboratory experiments, dispersion compensation succeeded. The fringe visibility was 0.93 for wide-band, where the central wavelength of light was 700nm, and bandwidth was 200nm, while 0.95 with a He-Ne laser. We used BK7 glass wedge for dispersion compensation. About fiber injection module, basic optical design has completed. The results of our fiber interferometer could contribute to OHANA (Optical Hawaiian Array for Nanoradian Astronomy) project. We present new science targets, white dwarves and T Tauri stars, and an 800 m delayline concept in CFHT for the project.

The next generation of stellar interferometer arrays will develop methods for observing fainter sources with greater resolution and create synthesized images at optical and IR wavelengths like those obtained from radio telescope arrays. Techniques using single-mode (SM) fiber optics offer significant advantages for future ground and space-based interferometers. SM fibers and integrated optics components can be used for nearly lossless transport and combination of light beams in a stellar interferometer, avoiding the multiple lossy reflective surfaces that must be kept in precise alignment with conventional optics. Furthermore, SM fibers spatially filter light corrupted by atmospheric seeing fluctuations or optical aberrations, increasing the fringe visibility and potentially improving measurement accuracy for bright sources by an order of magnitude. Controlling the dispersion and polarization properties of SM fibers is possible, but the poor coupling efficiency to aberrated light is a major limitation. MC fibers consist of a symmetrical arrangement of SM fiber cores inside a common cladding. One MC fiber is placed at the focus of each telescope, where light couples into the fiber and propagates through the cores for a short distance. Each MC fiber is then drawn apart into individual single-core fibers, and the resulting SM fiber beams are interfered pair-wise with beams from other telescopes. MC fibers are predicted to have an improved coupling efficiency over standard SM fibers, and the MC fiber geometry is well suited for transporting and combining light beams with minimal losses in an interferometer array. Computer simulations of fiber-linked interferometer arrays were performed to evaluate the performance of MC and standard SM fibers with different conditions of atmospheric, photon and detector noise. The effects of waveguide and material dispersion over a broad band at visible wavelengths are also included. The simulations determine the fiber modes, calculate the light coupling, propagate light through the fibers, and measure the beam correlations. Photon and detector noise are added and noisy estimates of the fringe power and bispectrum are found for the interferometer baselines. Statistics are then calculated over large ensembles and the measurements are processed to reconstruct an image of the source object. MC fibers are found to have greatly improved coupling efficiency over conventional SM fibers to aberrated light, and the noise sensitivity of visibility measurements and images also improves under certain conditions. Simulated images are shown and attempts to make MC fibers are discussed.

Several scientific topics linked to the observation of extended structures around astrophysical sources (dust torus around AGN, disks around young stars, envelopes around AGBs) require imaging capability with milli-arcsecond spatial resolution. The current VLTI instruments, AMBER and MIDI, will provide in the coming months the required high angular resolution, yet without actual imaging. As a rule of thumb, the image quality accessible with an optical interferometer is directly related to the number of telescopes used simultaneously: the more the apertures, the better and the faster the reconstruction of the image. We propose an instrument concept to achieve interferometric combination of N telescopes (4 ≤ N ≤ 8) thanks to planar optics technology: 4 x 8-m telescopes in the short term and/or 8 x 1.8-m telescopes in the long term. The foreseen image reconstruction quality in the visible and/or in the near infrared will be equivalent to the one achieved with millimeter radio interferometers. Achievable spatial resolution will be better than the one foreseen with ALMA. This instrument would be able to acquire routinely 1 mas resolution images. A 13 to 20 magnitude sensitivity in spectral ranges from 0.6 to 2.5 μm is expected depending on the choice of the phase referencing guide source. High dynamic range, even on faint objects, is achievable thanks to the high accuracy provided by integrated optics for visibility amplitude and phase measurements. Based on recent validations of integrated optics presented here an imaging instrument concept can be proposed. The results obtained using the VLTI facilities give a demonstration of the potential of the proposed technique.

We describe a preliminary experimental study of an interferometer built with two 500-meters-long arms made of polarization maintaining optical fibers. The control of the field polarization state along the single-mode fiber arms enables to measure fringe contrast up to 93% with a laser source emitting a 1290nm carrier wavelength. Preliminary contrast measurements achieved with broadband spectrum sources exhibit differential dispersion effect resulting from fiber inhomogeneities. Partial compensation of this effect is achieved by introducing additional fiber pieces on one arm. Moreover, we experimentally characterize the differential chromatic dispersion evolution as a function of the various additional fiber sections. Using the channeled spectrum method, a spectral analysis of the interferometric mixing allows to accurately measure the differential effect of chromatic dispersion i.e. second and third order term of the spectral phase shift.

Interferometry beam combiners that use optical waveguides, i.e. optical fibers or integrated optics, become popular in optical interferometry because of their flexibility, but also in the case of single-mode waveguide because of their properties of spatial filtering that increases the accuracy of interferometric measurement in an atmosphere-perturbed environment. However we know very little about the way the electric field propagates and even less about the correlation between the different beams of an optical interferometer. In this paper, we present in this paper an analysis of single mode optical waveguides in the framework of stellar interferometry. We first analyze the output electric field using radiated modes and show that the rejection rate we can derive in the case of nulling interferometry depends on many parameters, including the flux integration radius and the force of the aberrations. Secondly, since the interferometric equation can be interpreted in terms of carrying wave that carries respectively the optical power received by each telescope and the coherent power between two telescopes, we show that the interferometric equation involves a quantity called the modal visibilities which is not equal to the object visibility. The relationship between the two visibilities and the behaviour in the presence of atmosphere are also presented.

Following the first demonstration of an all guided two-beam stellar interferometer designed for space missions, we report an experiment recombining the beams coming from three telescopes using only guided optics or integrated optics components. This additional aperture could give us the possibility to achieve an image reconstruction using the phase-closure technique. We focus these calibration experiments on the interference fringe contrast measurements and the evolution of the phase-closure term versus the differential dispersion effects induced by the stretching of fiber delay lines.

This paper reports numerical simulations of spatial filtering using optical fibers or integrated optics components. The filtering capability of the waveguide is evaluated by launching a non-guided optical field and measuring the power rejection along the propagation, over a given integration area. The Beam Propagation Method (BPM) allows to compute the transverse field distribution all along the guide, whose refractive index profile has been defined beforehand. This work is focused on the study of straight waveguides.

The photonic crystal fibers are microstructured waveguides currently developed in the frame of fiber telecommunication. This study is mainly focused on the improvement of dispersion property and wide spectral single-mode operating domain. In the astronomical context, this kind of fiber is a good candidate to design a fiber linked version of a stellar interferometer. This paper gives an overview of the properties of these new waveguides and evaluates their potential in the frame of stellar interferometry.

We present preliminary results of injection tests in single mode fibers, conducted in August 2002 at Canada-France-Hawaii Telescope. Single mode fibers for J, H and K band where placed at the focus of the adaptive optics PUEO. The goal of this experiment was to prepare the next step of the OHANA project: an interferometric fiber link between two telescopes on top of Mauna Kea.

Spatial filtering is a critical issue to achieve nulling interferometry in the framework of spatial missions aimed at the detection of exoplanets. Several working interferometric instruments take benefit of guided optics for spatial filtering in the near infrared wavelengths and thus provide accurate visibility measurements. Furthermore planar optics would also provide beam combining functions within a single compact and stable device. Existing telecom technology allows beam combiner manufacturing for 0.8-1.8 micrometers wavelengths. Adaptation of these technologies is required to cover the scientific domain of ground based interferometry in each atmospheric spectral band and of the spatial missions like IRSI/DARWIN and TPF dedicated to thermal infrared wavelengths [4-20 micrometers]. We present here some of the most promising materials and their associated technologies for the thermal infrared range. For each of these solutions, based on chalcogenide glasses, semiconductor materials and hollow waveguides, we present some manufactured components with their optical characterizations. We also present a method and test-benches to measure the single-mode wavelength range of waveguides.

The phase closure is well known to remove the phase biases in a three-arm interferometer. This property is very useful to avoid the phase effect of the atmosphere in stellar interferometry. In this paper we theoretically investigate the effect of differential dispersion in a three-arm interferometer. We demonstrate that phase closure is corrupted by this spectral behavior. This theoretical analysis is illustrated by experimental results on a fiber version of a stellar interferometer. The good accordance between the numerical simulations and the experimental results valid this model.

Integrated optics technologies are an attractive alternative to classical bulk optics for the beam combination function of an interferometer. We propose a new integrated optics combiner for three apertures giving access to the closure phase on each output. It uses a multimode interference combination scheme realized by ion exchange on a glass substrate. This paper describes the theoretical behaviour of the beam combiner and its design constraints. Its interferometric behaviour is simulated and first experimental results using for the first time, as far as we know, a Near field Scanning Optical Measurement (NSOM) technique are discussed.

We discuss the potential of interferometric studies of nearby galactic nuclei with long-baseline interferometric facilities. Information on the morphology of galactic centers has so far been limited to angular sizes corresponding to the diffraction limit of 6-10 m class telescopes. Optical and near-infrared interferometry could in principle be used to reach significantly higher angular resolution, but has so far only been used for bright objects due to the small collecting areas of existing interferometers. Right now, the first interferometers consisting of 8-10 m class telescopes are starting operations and, hence, will soon allow us for the first time to study galactic centers on angular scales which are of an order of magnitude smaller than ever before, i.e. on scales corresponding to baselines of up to 100 m. We discuss these facilities and report on the observational techniques and strategies which are relevant for interferometric observations of these objects. We review imaging results of nearby galactic centers with highest angular resolution so far, with an emphasis on our bispectrum speckle interferometry studies of the core of the Seyfert galaxy NGC 1068. Employing these results, we analyze how near-infrared interferometry can discriminate between the different scenarios which are consistent with our current knowledge based on observations. In particular, characteristic sizes of the circumnuclear dusty torus can be derived with higher precision, additional dust components and the inner part of the jet can be identified, and radiative transfer models of the torus can be better constrained. Furthermore, the flux contribution of central source components can be separated from those of the torus, and thus they can be modeled in more detail. These investigations may ultimately result in a refinement of the unification scheme of galactic nuclei.

The Optical Hawaiian Array for Nanoradian Astronomy (OHANA) will allow resolutions on the order of a fraction of a milliarcsecond in the near infrared. This corresponds to the suspected size of the Broad Line Region and might even allow to study the structure of the base of jets. Preliminary studies with baselines on the order of or greater than 200 meters will be needed to understand these complex objects and successively refine existing models. The continuum visibilities will distinguish any unresolved source from resolved ones. Spectrally resolved visibilities in the H band and in Paschen lines will directly test competing models for the velocity structures that produce the observed line broadening. According to the unified model, the results of these measurements may be expected to depend on the relative viewing geometry. In Phase II, OHANA will be able to select, from many dozens of candidate sources, those most likely to present distinctive differences. A preparatory survey with adaptive optics is already under way. Four different scientific cases involving the OHANA array are reviewed: Active Galactic Nuclei geometries, Broad Line Region tomography, Broad Line Region dynamics, absolute distance measurements with Active Galactic Nuclei.

Simultaneously mapping the abundance inhomogeneities and the magnetic fields of chemically peculiar (CP) stars is essential to improve our understanding of stellar magnetism and its key role in structuring stellar atmospheres, in particular relative to ion migration and chemical stratification. However, magnetic fields and chemical inhomogenities tend to have similar effects on classical observables. Magnetic and abundance maps have therefore to be reconstructed most often either independently or in making a priori assumptions. To overcome these difficulties, we propose to take benefit of optical aperture synthesis arrays to resolve local magnetic structures and patchy stellar surfaces. This requires ability to resolve polarimetrically magnetically-sensitive spectral lines, and thus to add a polarimetric device at the combined focus of an interferometric array. Within this instrumental context, it becomes possible to map magnetic fields with visibility and phase measurements in circularly polarized light and to map the chemical inhomogeneities thanks to "classical" interferometric measurements (i.e. without the polarimeter). In this paper, we show that the interference fringe phase is the suitable observable for polarimetric measurements and for mapping patchy surfaces (see also Jankov et al. in these proceedings). We present some illustrative cases of different magnetic topologies and abundance distributions. We focus on two well-known CP stars, βCrB and α²CVn, and we show observational predictions with different instruments currently in operation (GI2T, VLTI).

OHANA will be a near-infrared long-baseline interferometer. It will be located at the summit of Mauna Kea and will link, by fiber optics, the existing telescopes equipped with adaptive optics into a giant interferometer. OHANA will have baselines up to 4 times longer than VLTI's. The improved resolution will be complementary to VLTI's and will be very well suited to study the star formation processes. Indeed, measuring and understanding the circumstellar environment of young stellar objects (YSOs), especially their accretion disk, is the key to understand the formation of planetary systems. Up to an age of about 10 million years, these disks are thought to be rich enough in dust and gas to host planet formation in their central regions. But this affirmation relies solely on an extrapolation of measurements and models of the outer, tenuous and cold parts of disks where planets cannot form because the measured density is too low. One can therefore rightly wonder how realistic these extrapolations to the central regions are and how secure is the claim that planets are forming around the young solar-like T Tauri stars. The disks' central region, located within a few stellar radii from the center, will become observable by OHANA. In this contribution we will show that OHANA will allow to measure and understand the structure of the accretion disk and differentiate between different models: equatorial vs. magnetospheric accretion; magnetized vs. standard disks, etc. These observations are fundamental to understand the link between the accretion process and the outflow/jet phenomenon frequently observed in these stars.

In this poster, we examine the science potential of an 800 meter interferometer such as the OHANA Array. The working assumptions are a K = 12 limiting magnitude, a 0.5 milliarcsecond resolution at K band, and a small (diffraction limit of individual telescope) field of view. The science cases described herein are by no means exhaustive and perhaps not even the ones that will eventually be carried out, but serve to illustrate the potential of the array. We expect that operation of the array will be proposal driven, so the actual science will come from the Mauna Kea communities. Our philosophy is that any measurement that can be made at a dedicated interferometer facility should not be a strong driver for OHANA. Therefore the science areas discussed in the poster focus on very high angular resolution measurements of faint sources. In some cases, science which can be addressed with simpler or dedicated facilities at an exploratory level can be carried to a significant new capability with OHANA. A limitng magnitude of 12 was obtained by simple computations, but first tests on the sky with the injection module (See adjacent poster on Phase I) will help narrow down this figure. At such sensitivity, Cepheid pulsations can be studied in considerable detail for a wide range of stellar parameters, leading to enhanced confidence in the accuracy of their use for distance measurement with minimal extrapolation or inferrence. The disk/star interaction zone in young stellar objects can be resolved with unprecedented detail for a range of masses and ages, providing direct information about the jet formation region, accretion rates and disk conditions. The broad line region of active galactic nuclei can be studied in a large number of sources of differing characteristics, testing specific models for AGN nuclear structure. For OHANA Phase III, a dual-star phase tracking capability is planned. With the resulting increased sensitivity, direct brown dwarf diameter measurement will provide a strong check on evolution models. Microlensing events could be resolved and provide unique new information about the lensing and the lensed objects.

The VLTI Calibrators Program is a common project between ESO and NEVEC. The main goal is to establish a network of measurements of calibrator objects with an accuracy high enough to fully exploit the different VLTI instruments. We started this project in 2001 by defining a list of objects to be used during the observations with the commissioning instrument VINCI. During the first year of observation (18th March 2001 - 18th March 2002), a total of 5060 observations have been recorded on 156 astronomical objects. More than 60% of the observations have been done on 63 calibrator objects. These calibrator data are currently analyzed to refine the measurements of the adopted diameters. After a brief description of the instrument and of the data reduction process, we describe the criteria used to establish a list of calibrators suitable for the commissioning instrument VINCI with baselines of up to 200m. We define a strategy to observe and analyze the data for the commissioning of the VLTI and of several baselines. We emphasize the difficulties of instrumental calibration to an accuracy of a few 0.1% and the necessity of a long term effort.

Three-dimensional radiation hydrodynamics simulation of the convective envelope and the atmosphere of a red supergiant (e.g. α Ori) have been performed, including ionization effects and realistic (grey) opacities. Only a handful of giant convection cells with lifetimes up to a few years are found in the envelope. The stellar surface itself is covered with smaller short-lived cells related to the well-known solar granulation but differing in many ways. Pressure fluctuations, acoustic waves, and shocks play an important role affecting the surface energy fluxes and accordingly the emitted intensity. The interaction with convection and waves generates observable large-scale high-contrast surface features. The complexity of the surface structures in the numerical models which will probably be even higher in nature calls for interferometric observations with high spatial and some temporal resolution. The comparison with the results of numerical simulations will allow to deduce information about the nature of the surface phenomena and fundamental stellar parameters.

In January of 2002, we began routine observations using the Navy Prototype Optical Interferometer with six siderostats operating simultaneously. We present recent images obtained with this 6-beam operations mode of the NPOI. We report on the implemented system enhancements that make this possible, such as the beam combiner, fringe detection electronics, fringe tracking algorithm and the control system. We discuss the present state of the NPOI data analysis, including amplitude and phase bias corrections, cross talk considerations, and calibration issues.

The first-generation COAST array is now primarily operated as a tool for astrophysics, with any development work aimed at improving observing efficiency and at prototyping hardware for future arrays. In this paper we summarize the full range of astrophysical results obtained with COAST in the previous two years. Results of a program to investigate hotspots on red supergiant stars are presented in detail.

As the number of optical interferometers increase, multi-facility observations become both feasible and scientifically interesting. For imaging of complex sources, the capability of increasing (u,v) coverage by using multiple arrays may be necessary for accurately interpreting the fringe visibility and closure phase data. Toward this end, coordinated observations with the IOTA interferometer and Keck aperture masking have been carried out to test techniques for synthesizing images using data from heterogeneous arrays with sparse (u,v) coverage. In particular, we will focus on how the image prior in the Maximum Entropy Method can be used to efficiently incorporate very high spatial frequency information with "low-resolution" data for imaging the generic prototype "Star + Dust Shell" image morphology. Preliminary results using real data for a few dusty evolved stars are presented.

The U.C. Berkeley Infrared Spatial Interferometer (ISI), previously described by Hale et al. as a two-telescope stellar interferometer operating in the mid-infrared regime of 9-12 μm, is now testing a recently constructed third telescope and centralized laser local oscillator and beam combining facilities for operation of a three-telescope system with phase closure. This new system will allow the ISI Array to make measurements of stellar photospheric and dust asymmetries in the mid-IR, measurements which have previously been lacking. The new facilities and instrumentation required for three-way heterodyne beam combination and fringe detection, together with techniques for measuring the closure phase, are described. Both hardware and software methods of beam combination are used for producing three simultaneous fringe visibilities and a closure phase measurement. Additionally, the phase closure signal-to-noise of pairwise and of all-in-one beam combination are contrasted.

In optical synthesis imaging, the incomplete sampling of the complex visibility function, atmospherically induced phase perturbations, detector noise, and low light levels all limit one's ability to produce high-fidelity images. One improves upon these limitations by using iterative non-linear deconvolution and self-calibration techniques to produce image models that are consistent with the data. The question of image fidelity, or how well these models faithfully represent the true object, is an important question to ask. A possible answer could be formulated using the concepts of statistical information theory. We apply the concepts of Shannon Information to an optical interferometer and propose methods for monitoring the information content of images as a measure of image quality.

A simple desktop optical interferometer is described and demonstrated as a teaching tool for concepts of long-baseline stellar interferometry. The interferometer is compact, portable, and easily aligned. It sits on a base 8" x 10" and uses an aperture mask which is mounted to rotate within a precision ball-bearing. Fringes produced from an artificial star are observed through a microscope by means of a video camera and are displayed on an overhead television monitor. When the aperture mask is rotated rapidly, the rotating fringe patterns seen on the monitor are observed to synthesize sources that are unresolved by individual holes in the mask. Fringes from an artificial double star are used to illustrate the relationship between fringe visibility and source structure and to demonstrate image synthesis.

We report on the experimental validation of the photon noise correction applied on interferometric data, in the frame of an imaging laboratory test using the end-to-end demonstrator "Optical Aperture Synthesis Technologies 2". The light flux is scaled to astronomy level to achieve a quantitative comparison between the contrasts and phase closure measurements simultaneously measured in photon counting regime and in high flux level regime. This experiment uses a three-telescope array linked by optical fiber to the mixing station. We show that it is possible to correct properly the interferometric data with light power levels reaching 1 femto-Watt at a 680 nm mean wavelength. This corresponds to a photon flux close to 2.5 photons per millisecond and to 25 photons for each snap shot.

The use of single-mode optical fibers in interferometers allows both easy transport of the telescope beams and very accurate visibility measurements, thanks to their spatial filtering property. Unfortunately, spatial filtering reduces the field of view coupled in the interferometer to approximately the size of the diffraction limit of the largest telescope in the array. More seriously, the time-variable wavefront aberrations induced by the atmosphere continuously change the coupling properties of the fibers, and consequently, what each arm of the interferometer "sees" on the sky. These field of view effects are especially important for the next generation of interferometers, such as OHANA (Optical Hawaiian Array for Nanoradian Astronomy), where adaptive optics makes it possible to use large individual telescopes, and the increased sensitivity makes possible the observation of fainter, more extended objects with complex morphology. In this paper, the impact of these effects is studied for observations of a variety of sources by OHANA. It is shown that the errors in visibility measurements can be reduced by signal processing techniques. Possible solutions to extend the accessible field of view of fiber-based interferometers are also presented.

This paper presents a theoretical limit to the field/resolution ratio for the imaging mode of aperture synthesis interferometers. This limit is a function of both the number of apertures in the interferometric array and the dynamic range of the visibility and phase data. It does not depend on the optical setup of the instrument. This work is based on the theory of information.

Wide Field Imaging is a natural extension to single boresight interferometry with an optical/infrared telescope array. It is an important tool to obtain interferometric data of extended objects and for astrometric measurements. Visibilities from many points on the sky can be obtained in one shot saving observation time. In this paper we introduce a new technique, the "staircase mirror" concept which offers potential advantages with respect to existing techniques to perform wide field of view operation on an optical stellar interferometer. This new concept is based on a pupil plane recombination scheme with an automatic field-dependent path length compensation in the intermediate image field of each array single telescope. The pathlength compensation is obtained via a staircase mirror whose position and step depth are a function of the pointing direction and the baseline vector of each telescope. The mirror must be actuated to follow the changes of the baseline projected on the entrance pupil of the telescope. We have calculated the differential delay as a function of the field angle, studied and designed an experimental setup to show the applicability of the method and performed simulations for the Very Large Telescope Interferometer.

The Delft Testbed Interferometer (DTI) will be presented. The main purpose for the DTI is to demonstrate the feasibility of homothetic mapping, both fixed and under scanning conditions. The driving design issues behind the DTI will be presented together with a list of experiments to be conducted with the DTI system in the field of wide field imaging.

The field of view for the first generation VLTI instruments (VINCI, MIDI, AMBER etc.) is limited to the diffraction limited beam of a single telescope (for a 8.2-meter telescope at 2-micron this is 0.6 arcseconds). However, the VLTI infrastructure with its main delay lines, transfers a 2 arcseconds beam from the telescopes to the interferometric laboratory. When we discuss wide field imaging in the context of this paper, we refer to these 2 arcseconds. Although most current optical interferometers use Michelson pupil plane beam-combination there is a convincing scientific justification for wider field imaging capabilities to allow astrometry, photometry and ultimately spectroscopy of crowded fields (e.g. galactic center, globular clusters), binaries, stellar disks and dust shells. In this paper we present a newly developed tool to model a fringe tracker for a Fizeau interferometer by retrieving the piston of each individual telescope from the combined point spread function. This study shows that the pistons can be retrieved. Detecting the signal in at least two wavelength channels with an energy sensitive detector (e.g. Superconducting Tunneling Junction) allows compensation for the atmospheric piston and successively coherent integration of a science channel. The examples provided in this paper are compliant with the VLTI infrastructure and are part of a larger project to study the feasibility of a wide field imager for the VLTI as a third generation instrument. Finally we present a conceptional design for a wide field imaging for the VLTI, referred to as VLTI-WIDY.

This presentation reports the status of our study concerning the imaging properties of the Large Binocular Telescope (LBT) interferometer, and namely the effect of limited angular coverage and partial adaptive optics (AO) correction. The limitation in angular coverage, together with the correlated problem of angular smearing due to time-averaging of the interferometric images, is investigated for relevant cases depending on the declination of the observed object. Results are encouraging even in case of incomplete coverage. Partial AO-correction can result in a wide range of image quality, but can also create significant differences within a same field-of-view, especially between a suitable reference star to be used for post-observation multiple deconvolution and the observed object. Our study deals with both the problem of space-variance of the AO-corrected point-spread function, and that of global quality of the AO-correction. Uniformity, rather than global quality, is found to be the key-problem. After considering the single-conjugate AO case, we reach to some conclusions for the more interesting, and actually wide-field, case implying multi-conjugate AO. The whole study is performed on different types of object, from binary stars to diffuse objects, and a combined one with a high-dynamic range.

Fizeau interferometry at the Large Binocular Telescope (LBT) offers significant advantages over other facilities in terms of spatial resolution, field of view, and sensitivity. We provide an update of the LINC-NIRVANA project, which aims to bring a near-infrared and visible wavelength Fizeau beam combiner to the LBT by late 2005. As with any complex instrument, a number of detailed requirements drive the final design adopted.

The development of optical interferometry will literally open a new dimension in the study of binary stars. As a prelude to the session, this paper reviews those properties of stars in binary systems that are of greatest scientific interest and summarizes what has been learned with the traditional tools of photometry and spectroscopy. We then discuss the information provided by interferometry and how it is most powerfully exploited in combination with other, complementary tools. This is illustrated by a few recent cases in which optical interferometry has helped to break new ground in ways that point towards the future. Finally, some recommended procedures are listed which may help to optimize the scientific returns of interferometric studies of binary stars.

Georgia State University's Center for High Angular Resolution Astronomy (CHARA) operates a multi-telescope, long-baseline, optical/infrared interferometric array on Mt. Wilson, California. Since its inception, one of the primary scientific goals for the CHARA Array has been the resolution of spectroscopic binary stars, which offer tremendous potential for the determination of fundamental parameters for stars (masses, luminosities, radii and effective temperatures). A new bibliographic catalog of spectroscopic binary orbits, including a calculated estimate of the anticipated angular separation of the components, has been produced as an input catalog in planning observations with the Array. We briefly describe that catalog, which will be made available to the community on the Internet, prior to discussing observations obtained with our 330-m baseline during the fall of 2001 of the double-lined spectroscopic systems β Aur and β Tri. We also describe the initial results of an inspection of the extrasolar planetary system υ And.

ESO is embarking on the construction of a complex and high performance dual feed system that will allow Phase Referenced Imaging and Microarcsecond Astrometry (PRIMA) on the VLT Interferometer on Cerro Paranal in Chile. In this paper, I will describe in some detail the scientific objectives of this facility that drive the technical specifications and justify the chosen priorities. Of particular importance because of its uniqueness and rich variety of scientific applications will be the early development of the components allowing very high precision astrometry on sources such as extrasolar planets, binaries in nearby clusters, microlenses in the halo, and stars in the circumnuclear cluster in the galactic center.

With the imaging and astrometric facilities coming up as the next major leaps forward in interferometry at the VLT we examine options for beam combination of multiple beams, both in the Science Channel and in the Reference Channel. We look at the science we expect to become feasible as soon as phase referencing becomes available, and the hopes for the future of this facility with the advent of more ATs and next-generation beam combiners. Contrary to "conventional wisdom," the fringe tracking accuracy for a multi-element array is not worse than for a single-baseline interferometer with equal telescope size. We discuss applications of phase referencing to observations of faint targets, and a double-differential phase technique for spectroscopy of extrasolar planets. We point out that phase-referenced methods have some important advantages compared with the closure-phase technique for imaging of extended sources.

A fixed delay interferometer combined with a post-disperser is a new technique for high precision radial velocity (RV) measurements. The Doppler measurements are conducted by monitoring the stellar fringe phase shifts of the interferometer instead of absorption line centroid shifts as in the echelle. High Doppler sensitivity is achieved through optimizing the optical delay in the interferometer and measuring multiple fringes over a broadband. The broadband operation is achieved by using the post-disperser for dispersing fringes in different wavelengths. Comparing to the state-of-the-art cross-dispersed echelle spectroscopy, this interferometer technique provides almost identical RV precision based on photon statistics. However, the interferometer method has a potential for lower systematic noise due to its simpler instrument response than the echelle. The interferometer can be optimized for higher throughput than the echelle. The interferometer approach also allows fringes to be recorded in one dispersion order instead of many cross-dispersed echelle orders. Therefore, this instrument opens up a great opportunity for multi-object observations to allow all sky surveys for extra-solar planets at moderate sized wide field telescopes. Initial observations with a prototype at the Hobby-Eberly 9 m and Palomar 5 m telescopes demonstrate ~9 m/s Doppler RV precision with stellar fringe data recorded on a 1kx1k CCD detector (or 140 Å wavelength coverage), a S/N ~ 120 per pixel and a post-disperser spectral resolving power of R = 6,700. This precision is consistent with the photon noise limit. Future improvement in wavelength coverage and wavelength calibration can reduce the Doppler error to a few m/s or less.

Up until now and with the first generation of the next optical interferometers, only a few telescopes, such as 3 or 4, are enabled for gathering the starlight. In such a case, short night-time observational runs will only provide small number of measured baselines. The lack of spatial frequencies mapped in the uv plane thus prevents from a reconstruction of the studied object. It therefore implies to define an a-priori model and to constraint its parameters from the available measurements. We propose here to fit the model parameters from a least square minimization-type approach. We develop a general formalism that can be applied to the three common interferometric observables. These observables are the square visibility, the closure phase and the differential phase. We also pay attention on the formal analysis of the error on the estimated observables and on the resulting error on the model parameters. This approach allows us to run realistic pre-observational simulations, to estimate the performances of the telescope/recombiner instrument, and to sense the feasibility of the intended observation. We finally apply this theoretical study to the recombiner AMBER on the VLTI when pointing at a single star.

Light is not a scalar wave. We only get away with treating it as such when the degree of polarization is very low. This condition often holds for seeing-limited single telescopes, but becomes less likely at spatial resolutions typical for interferometers. For the interferometric environment, optical polarimetry may need to assimilate radio-polarimetric concepts. In particular, the Stokes parameters should be defined in terms of complex correlations rather than as differences of orthogonally-polarized fluxes. Polarization effects in the Coudé train and delay lines spoil the accuracy of traditional quasi-scalar interferometers. An alternative optical architecture is proposed, using traditional (i.e. single-beam) optical polarimetry in the correlator, but 'radio-type' transfer of light from telescope foci to correlator (i.e. 2 clean, fully-polarized, signals from each telescope). Such a fundamental solution can eliminate errors due to inclined mirrors (phase shifts and added polarization). The architecture enables full-Stokes polarimetry at the resolution of the interferometer, but also a 'no-polarization-desired' mode which does not necessarily involve loss of signal-to-noise ratio and yet is free from polarization-induced errors of photometry. Existing polarization components permit a very wide instantaneous bandwidth (e.g. 0.3 to > 1 μm, matching CCD or STJ detectors).

The calibration process of long baseline stellar interferometers requires the use of reference stars with accurately determined angular diameters. We present a catalog of 374 carefully chosen stars among the all-sky network of infrared sources provided by Ref. 1. The catalog benefits from a very good sky coverage and a median formal error on the angular diameters of only 1.2%. Besides, its groups together in a homogeneous handy set stellar coordinates, uniform and limb-darkened angular diameters, photometric measurements, and other parameters relevant to optical interferometry. In this paper, we describe the selection criteria applied to qualify stars as reference sources. Then, we discuss the catalog's statistical properties such as the sky coverage or the distributions of magnitudes and angular diameters. We study the number of available reference stars as a function of the baseline and the precision needed on the visibility measurements. Finally, we compare the angular diameters predicted in Ref. 1 with existing determinations in the literature, and find a very good agreement.

This paper deals with the experimentation and simulation of the telescope directivity effects in the frame of stellar interferometry. Contrasts and phase closure can be corrupted when using large aperture. The corruption becomes very critical if telescope pointing error occurs. We report an example on numerical simulation of contrast and phase closure corruption to be measured for 40 cm and 100 cm telescope diameters. The OAST2 test bench has been used to experimentally demonstrate this data corruption.

The study of Young Stellar Objects (YSOs) is one of the most exciting topics to be handled by long baseline optical interferometry. The magnitudes of these stars are at the edge of capabilities of current optical interferometers limiting the studies to a few of those but are well within the capability of coming large aperture interferometers like the VLT Interferometer, the Keck Interferometer, the Large Binocular Telescope or 'OHANA. The milli-arcsecond spatial resolution reached by interferometry allows to probe the very close environment of young stars down to a tenth of an astronomical unit. In this paper, I review the different aspects of star formation that can be tackled by interferometry: circumstellar disks, multiplicity, jets, etc. I present recent observations performed with operational infrared interferometers, IOTA, PTI and ISI, and I show why in the next future we will extend these studies with large aperture interferometers.

Dusty astrophysical systems, including young stellar objects, evolved late-type stars and colliding-wind binaries have all proved ready targets for interferometric imaging studies. New results from the established aperture masking project operating in the near-infrared on the Keck-I telescope are presented. A number of objects exhibit detailed morphologies and motions of circumstellar material which convey a great deal of astrophysical insight. Such results give a preview of the immense potential of the new generation of interferometers capable of going beyond the brightest few prototypical sources presented here.

The Navy Prototype Optical Interferometer has recently been equipped with specially-designed filters that pass Hα emission in a 2.5 nm band, suppress the continuum 50 nm to either side, and pass the continuum further from the Hα line. These filters allow fringe tracking on continuum light while taking data at Hα. Five- and six-aperture NPOI configurations have also been implemented recently. The improvement in U-V coverage with these configurations promises greater image fidelity in multi-spectral imaging as well as in specific lines, such as the very interesting Hα line. Using an array simulator operating in the AIPS++ environment, we simulate observations of Hα emission, assuming approximate source structure taken from earlier work in the literature. These simulations demonstrate the increased imaging capability of multi-aperture arrays and help define optimum Hα observation strategies.

An optical interferometer with long baselines and high spectral resolution could be used to determine the rotation rate of a star and the orientation of its rotation axis. Orientation is information that has previously been determined only statistically for single stars. Knowing the orientation would allow solution of currently intractable problems in star formation and stellar evolution.

Similarly as the technique of Doppler Imaging from spectroscopic observations, Differential Interferometry makes it possible to measure the disturbances of photocentroid location of an unresolved star as a function of wavelength and to deduce the corresponding stellar map. We show the imaging potential of a tomographic technique which combines time-resolved spectroscopy and long baseline interferometry, providing information that cannot be obtained otherwise with each of these techniques taken at once. In particular, here we consider the example of mapping abundance inhomogeneities, performing numerical experiments with realistic spectral resolutions and signal-to-noise ratios expected for operating (VLTI, GI2T) or close-to-operating long baseline interferometers (Chara, Keck). We show that the accurate maps of stellar surface abundance distribution can be obtained using regularized inversion by Maximum Entropy method. The technique is also applicable to other classes of stellar surface imaging as magnetic field and temperature spots but within the classical instrumental context (without polarimetric device) it can hardly discriminate among different distributions. We discuss the importance of Spectro-Polarimetric Interferometry observations (Rousselet-Perraut et al., this proceedings) in order to discriminate and simultaneously map abundance/temperature inhomogeneities and magnetic fields of chemically peculiar (CP) stars.

The field of nulling interferometry has seen significant progress over the past several years, in both the conceptual and experimental arenas. Deep, broadband nulling has been demonstrated at optical wavelengths, the techniques have seen initial implementation on telescopes, and the introduction of a symmetric beam-combiner concept has eliminated many of the residual obstacles. Here an overview is provided of promising techniques for effecting the deep cancellation of starlight, and recent results obtained with laboratory and astronomical nulling interferometers are discussed. The next step is the exploitation of nulling techniques at 8-10 m class separated-aperture telescope facilities, and in this vein, a brief overview of the architecture of the Keck Interferometer Nuller is also provided.

A wideband interferometer is sensitive to the effects of longitudinal dispersion which affect the interfering light beams unequally. At shorter wavelengths the major effect of dispersion is from dry air itself, while at mid infrared wavelengths the effect of water vapor is dominant. MIDI, the future 10 micron instrument of the VLTI, will experience significant effects from the imbalances in the water vapor content affecting the paths of the two interfering beams. This imbalance will include terms due to the unbalanced air paths in the delay line, random atmospheric humidity fluctuations in the lines-of-sight to the star, and random humidity variations inside of the VLTI delay line tunnels. Large amounts of dispersion, if not monitored, can reduce the accuracy of measured visibility amplitudes. Measurements of the visibility phase as a function of wavelength will be highly sensitive to dispersion. This will then become a source of noise in results dependent on the phase of interference, such as imaging of non-symmetric objects, or detection of faint companions. In addition to dispersion over the 7 - 14 micron region detected by MIDI, observations using phase tracking with detection in the near IR, could be catastrophically affected by differential phase delays between the 2 micron and 10 micron bands. Dispersion measurements from VINCI observing in the K band, both due to dry air and to water vapor, are presented. Combining VINCI results with published data from millimeter wave measurements leads us to expect atmospheric differential water vapor fluctuations to exceed 1 mole/m² rms over typical baselines. Specific effects from such a level of unmonitored dispersion variations are presented, which demand corrective action. Various solutions to monitor water vapor dispersion in realtime are considered.

The control of longitudinal dispersion, which determines the position of the null fringe as a function of wavelength, is central to the problem of producing deep broadband interferometric nulls. The dispersion is the sum of terms due to environmental factors such as the dry-air and water-vapor atmospheric seeing, the unbalanced air column due to the unequal delay-line paths between the telescopes the combiner, and to the distance from the central fringe. The difference between an achromatic nuller and a normal constructive combiner operating at its first (chromatic) null can be thought of as an added longitudinal dispersion, which for the case of the Keck Interferometer is smaller than the environmental terms. We demonstrate that the sum of these effects can be adequately compensated by an appropriate thickness of ZnSe combined with an additional achromatic pathlength. The Keck Interferometer nulling combiners take advantage of this result. They are intrinsically constructive combiners made to produce achromatic nulls by inserting a ZnSe dispersion corrector into each of the four input beams. We describe the null fringe stabilization control algorithm and present calculations of the required loop bandwidth and precision. A potentially important advantage of the present approach is that the system will be able to function as either a destructive or constructive combiner, depending on the value of a single control-loop parameter (the target fringe phase).

Atmospheric turbulence is a major source of noise in any fiber-linked interferometer through the piston effect between the two arms of the interferometer which induces an erratic movement of the fringes. Because the stellar light has to be permanently cancelled by the central dark fringe, a ground-based nulling interferometer is not possible without stringent optical path delay (OPD) control. In this paper, we investigate the influence of the residual OPD error of a fringe tracking unit on the performances of a nulling interferometer. The accuracy required for this control system strongly depends on the observation wavelength. If we want to detect exozodiacal clouds ten times as dense as our zodiacal dust cloud in the near-infrared (L' band), the performances of the fringe tracker are crucial: the residual OPD should be about 20 nanometers RMS. This specification can only be reached if the fringe sensor is optimized for bright sources. In the mid-infrared (N band), the requirements are strongly relaxed: OPD control with an accuracy of 400 nm RMS is sufficient to be background-limited, but a lower residual OPD (about 50 nm RMS) is strongly recommended to reduce the stellar leakage.

Nulling stellar interferometry may enable the discovery of earth-like planets around other stars. In nulling mode, the zero order fringe is destructive and on axis, thus cancelling light from a bright source and allowing detection of dimer off-axis features. To create a deep on-axis null, the phase must be shifted half a wave achromatically over a broad band. The phase shift is created by adding optical path thickness with dielectric plates. Plates of different materials can balance dispersion. The nulling solutions found for TPF (infrared) and SIM (visible) are promising. This paper presents the final results of a dissertation that developed a nulling beam combiner testbed. The deepest null achieved over the spectral region of 600 to 800 nm was 7x10^-3. The test bed revealed the extreme challenges of this technique and provided very valuable lessons to enable further implementations. The testbed first achromatized the null by actively controlling the optical thicknesses of the plates. The phase as a function of wavelength was measured by PSI on a spectrally dispersed fringe. The phase was fit to a model to determine the optical thicknesses. The eigenfunctions of the model were nearly collinear and consequently the dynamic range required of the phase data was very high and not supported by the hardware. The testbed then searched for the null fringe and locked on the null using a 300 Hz servo loop and on a grey fringe. The OPD was stabilized to 6 nm peak-to-valley.

There is great interest in techniques that potentially enable the direct detection of light from planets orbiting nearby stars. In one technique, nulling interferometry, the light from two or more separate telescopes can be destructively interfered to cancel the light from the star, allowing detection of the faint off-axis light from an orbiting planet or circumstellar disc. At the Jet Propulsion Laboratory, broadband white light nulls limited by dispersion effects have been previously reported. In this paper, we investigate other limiting effects using a monochromatic laser source. We report transient laser nulls as deep as 1 part in 10⁶ at a wavelength of 633 nm in a nulling interferometer operated in air. We discuss the techniques used to obtain these nulls and the residual sources of stray light and other null limitations.

Spatial or modal filters are essential parts of highly rejecting nulling interferometers. We review the principle of operation of both types of filters and explain the fundamental physical difference. We point out the filter's individual properties and potentials, and analyze practical limitations. For modal filters we discuss implementation alternatives, also with regard to their suitability for mid-infrared operation. For a single-mode fibre filter we analyze the broadband performance and the minimum length ensuring a prescribed filter action. We further present simulation results of a DARWIN-representative nulling interferometer breadboard which confirm the distinct improvement in rejection ratio due to spatial or modal filtering.

A tabletop rotational-shearing interferometer experiment has been constructed and operated at JPL to serve as a testbed for the mid-infrared (~10 μm) nulling beam combiners on the Keck Interferometer and the Terrestrial Planet Finder. The testbed is a pupil-plane combiner in which destructive combination of the incoming wavefronts is achieved using a rooftop mirror system in which the polarization vector is flipped along the vertical axis on one arm and the horizontal axis on the other. The optical pathlength along one arm is adjustable using a linear stage driven by picomotor and piezoelectric actuators. The combined light is focussed onto a single-pixel LN₂-cooled HgCdTe detector. In order to provide adequate sensitivity in the presence of the very bright thermal emission from the room-temperature optics, the light source is modulated and the output is demodulated using a lock-in amplifier. The optical pathlength difference (OPD) is stabilized under computer control by slowly dithering the actuated arm and balancing the leakage signal on either side of the null. The system has produced a stabilized null depth of < 10^-4 using a diode laser source emitting at a wavelength of 9.2 μm, and transient nulls of 10^-2 with a broadband thermal IR source in a 6.4% optical bandpass.

The IRSI/DARWIN spatial interferometer of the European Space Agency (ESA) is aimed at detecting extrasolar planets. The high difference in flux emission between the star and the planet is tackled by using nulling interferometry as a coronographic method. By star light extinction, one can retrieve the planet signal, and thus have access to high resolution imaging by interferometric measurements. Critical technological solutions are to be developed in order to reach the high level performances of such instruments. This is the scope of the Multi-Aperture Imaging Interferometer (MAI²) breadboard developed by Alcatel Space in an ESA contract. The goal of this laboratory experiment, based on integrated optics (IO) beam combination, is to obtain stable rejection of a star signal at a level of 10⁶.

Darwin is one of the most challenging space projects ever considered by the European Space Agency (ESA). Its principal objectives are to detect Earth-like planets around nearby stars and to characterize their atmospheres. Darwin is conceived as a space "nulling interferometer" which makes use of on-axis destructive interferences to extinguish the stellar light while keeping the off-axis signal of the orbiting planet. Within the frame of the Darwin program, the European Space Agency (ESA) and the European Southern Observatory (ESO) intend to build a ground-based technology demonstrator called GENIE (Ground based European Nulling Interferometry Experiment). Such a ground-based demonstrator built around the Very Large Telescope Interferometer (VLTI) in Paranal will test some of the key technologies required for the Darwin Infrared Space Interferometer. It will demonstrate that nulling interferometry can be achieved in a broad mid-IR band as a precursor to the next phase of the Darwin program. The present paper will describe the objectives and the status of the project.

This work is funded by ESA under ESTEC/Contract No. 14827/00/NL/CK. Astrium Germany has been awarded this first ESA breadboarding towards nulling interferometry. Interferometric nulling devices are essential ingredients in the TPF and DARWIN missions for suppressing the star light by a factor of 10⁶ over a wide wavelength range in the mid infrared. The current DARWIN baseline concept comprises six telescopes. The coherent combination scheme in the nulling mode operation foresees three nulling assemblies in parallel. The breadboard serves to demonstrate the deep and stable null required for an operational instrument. The demonstrator operates in the near infrared to save costs but its principle is fully applicable to the mid infrared. The nulling device is based on an autobalancing Sagnac core offering just one critical beam combiner. Two different ways of achieving the required π phase shift are implemented: a) arbitrary phase shift by dispersive phase shifter plates b) phase shift of π using periscopes (image flip) The target simulator features two point sources of adjustable radiometry and angular separation, representing a strong star and a weak planet. In addition, the sources can be also used to simulate a double star for demonstrating the basic DARWIN imaging mode. The simulator can be operated in two styles, namely as wavefront dividing star/planet source and, alternatively, as an amplitude dividing source, providing highly symmetric wavefronts to both interferometer arms. Because of its representativity for the DARWIN situation, the latter mode is the preferred simulator for quantitative nulling experiments. The breadboard design has been finalized in January 2002 and verified by detailed simulations. The entire hardware has been manufactured by end of July. Currently, nulling and imaging measurements are in progress to validate the per-formance of the selected approach. The project is part of ESA´s technology preparatory program for DARWIN, paving the way for a collaborative ESA/ESO guest instrument at ESO's VLTI with scientific implications.

Geometric phase modulation is used to realize achromatic phase shift in nulling interferometer, where Fresnel rhombs are key components of the geometric phase modulator. Experimental results show high extinction ratio in visible region. Extension of our scheme to the infrared region is also discussed.

A new design for a wavefront sensor suitable for low-order low-light correction is shown. The hybrid modal sensor, the Nine Element (NE) sensor, is compared with a curvature sensor and quadcell under single aperture applications. The design of the NE sensor allows the use of readily-available array detectors. We discuss the optimization of the design to maximize its performance with respect to the number of Zernike polynomials to detect and optical parameters, using a simulated annealing technique. Numerical simulations show the good SNR response low-light levels, and indicate a reduction in wavefront variance from 6.41 rad² to 2.01 rad². The sensitivity to tip/tilt errors is demonstrated to be comparable to a quadcell. Successful closed feedback loop operation results in corrected Strehl ratios of over 0.5. Improvements and future work is discussed.

The Magdalena Ridge Observatory now being planned for a site within the Magdalena Mountains near Socorro, NM will have an optical/near infrared interferometric imaging array as its primary observing instrument. We are presently evaluating the use of array telescope apertures ranging from 1.2 to 2.0 m in diameter for this application. Telescopes in this size range are many times the size of Fried's coherence parameter r₀ anywhere within the .6 to 2.2 μm wavelength range of interest and to be useful for interferometry will require the use of adaptive optics (AO) techniques to restore and maintain the spatial coherence of the telescope pupils. We review some of the practical limitations related to the use of AO systems on interferometer telescopes and discuss the enhanced interferometric performance that might thereby be attained.

The outline and the performance of the wavefront tilt correction system for the Mitaka optical and InfraRed Array, MIRA-I.2, is reported. The tilt control system consists of quadrant detectors and PZT-driven tilt mirrors. These components are controlled by three computers connected via local area network system. The control loop frequency of 2660Hz is achieved and this high loop frequency enables the stable operation of the control system with a bandwidth of approximately 80Hz in open loop at 0dB. Preliminary results of one night observation of α Lyr showed that this system suppress the relative wavefront tilt of incoming lights to 0.46 arcsec while it is 1.9 arcsec without control. With an aperture size of 90 mm and the wavelength of 750 nm, this means the tilt control system improves the visibility loss to 9% while it is 89% without control.

The resolution of a conventional telescope is determined by the spatial extent of the collecting surface, usually the primary mirror. Astronomical interferometers achieve increased fine detail by using unit telescopes spaced over large distances to increase the spatial extent. The required wavefront quality places very tight tolerances on the unit telescopes and they should be designed with the prime goal of meeting the wavefront specification. The unit telescope must be optimized for the role of a beam compressor rather than attempting to modify a conventional design. Two alternative designs that minimize the number of reflections in the telescope will be considered, a crucial feature in obtaining the lowest possible wavefront error and maximizing throughput. The first, a siderostat has fixed imaging optics and a large steerable flat mirror to enable sky tracking. The second, an "Alt-Alt" system consists of two intersecting altitude axes in a "gyroscopic type" structure. A small flat lies at the intersection of the altitude axes to direct the starlight at a constant height and direction out of the telescope. The benefits and limitations of each are shown along with the key design issues that determine the most appropriate unit telescope for implementation in an interferometric telescope.

As part of the Very Large Telescope Interferometer (VLTI), ESO signed in June 1998 a contract with AMOS S.A. covering the design, manufacturing and testing of the Auxiliary Telescope System (ATS), including three movable auxiliary telescopes (AT) and the associated site equipment. The project entered into manufacturing phase in mid-1999. As a prime contractor, AMOS S.A. had to deliver complete and fully operational telescopes ready to plug on any of the thirty stations of the VLTI sub-array. This work required an overall system level approach involving a multi-skilled team. This paper depicts the progress status and presents the results of the main tests -- except optical testing -- performed on the telescopes in AMOS facilities. Those tests are the earliest verification of high-level requirement specifications such as optical path length (OPL) stability, pointing error, main axis dynamic response, performed on the telescope fully integrated.

Interferometric observations at 10 micron combine the difficulties of the relatively new interferometric techniques with the problems of overcoming the strong and highly variable thermal background which are typical of thermal infrared observations. In particular, the detector subsystem must comply to strict requirements in terms of stability, of read noise, and of read out speed. Here we present the results obtained during laboratory test of MIDI, the Mid-IR interferometric instrument for VLTI. We selected as detector for the MIDI instrument a Raytheon 320x240 IBC array. We will discuss some of the aspects of the foreseen operation of MIDI, and the methods adopted to implement those on our detector system. We will show our results on detector stability, on its performances (in particular Quantum efficiency and read-out noise), and on the reaction to high fluxes. By using the possibility of hardware windowing, frame times of the order of 2 ms can be reached. Finally, we will show the characteristics of the detector when used in interferometric mode during tests of the whole MIDI instrument with both monochromatic and broad band calibration sources.

Most current CCDs cannot be used as optical interferometric sensors because the high readout noise disguises the small signal. However, new low light level charge coupled devices (L3CCD) have a large on-chip gain which can allow a signal to be detected above the noisy readout amplifier. This gain has a statistical nature, meaning that the photon input cannot be predicted exactly. We investigate several techniques for photon prediction at different light levels, and demonstrate how this affects the noise on the signal. Accurate signal estimation can be achieved with very faint signals, up to about one photon per pixel per read. Above this, accuracy gradually decreases, though our signal-to-noise ratio is never worse than square root(2n). Optical interferometry requires detection of very faint signals, and the use of an L3CCD is found to allow reproduction of interferometric visibilities to high precision. Custom instrumentation used for control is also detailed.

The Cambridge DIMMWIT was developed for the site of COAST (the Cambridge Optical Aperture Synthesis Telescope), a prototype optical interferometer. Unlike other differential image motion monitors, this design is portable in order to carry out seeing campaigns at the site of any optical inteferometer. Of particular interest is the site of a second-generation interferometer proposed by the MRO (Magdalena Ridge Observatory) consortium. The DIMMWIT design has two objectives: to measure the Fried parameter r₀ and the speckle lifetime tau₀, and to be easily transportable. Here, we outline the theory of differential image motion, the design of the DIMMWIT, describe how turbulence parameters can be measured with COAST, and compare measurements of the seeing conditions made simultaneously by the monitor and COAST.

Previous measurements of atmospheric density fluctuations have shown that a substantial fraction of seeing fluctuations occur within 100 feet of the ground, and that the power spectrum of path length fluctuations through the atmosphere has a somewhat smaller slope than that predicted by the Kolmogorov-Taylor approximation. To provide some possibility of appreciable path-length corrections, the ISI has assembled a system capable of measuring temperature changes at fifteen foot intervals of heights up to 70 feet from the ground. Analysis of temperature measurements made under a variety of conditions confirms previous results concerning the decrease in the magnitude of the fluctuations with altitude near the ground: the rms magnitude of the temperature fluctuations at an elevation of 70 feet is, on average, 52% of the mean rms value at 9 feet. However, these new measurements made at point locations show a power spectrum close to the Kolmogorov-Taylor prediction at frequenices up to 1.0 Hz, for average wind speeds above 2 m/s. In addition, correlation analysis between sensors located at the same elevation but separated by a given distance shows up to 50% correlation out to separations as large as 24 meters with wind speeds of a few meters per second, and indicate that Taylor's approximation applies over spatial distances in the range of 24 - 85 meters, or on timescales as large as ten seconds, and perhaps as large as 14 or 15 seconds. This makes path length corrections possible by temperature measurements at nearby locations.

We have developed a new algorithm for simulating atmospheric phase screens that is extremely efficient and flexible. It retains the N log N efficiency of algorithms based on the FFT and is able to take advantage of the small amount of atmosphere actually sampled by highly dilute arrays such as NPOI. Additionally, there is flexibility in the technique to relax (or adopt) the traditional "frozen screen" approximation. We describe the basis for the algorithms and how the code has been structured. Timing estimates are developed and we show preliminary results from the code which exhibit the correct behavior of phase difference power spectra with baseline.