This paper and oral presentation will describe the technology studies, the testbeds, and the architecture studies that will enhance the understanding and viability of a Terrestrial Planet Finder Coronagraph. Topics to be described fall in two categories: technology development and coronagraph mission design. The focus of the paper will be explanation of the tasks, their organization and current status.

One possible implementation of an optical coronagraphic approach to finding exo-solar planets incorporates a large, monolithic primary mirror (PM) that is approximately 4 meters by 10 meters in size. The optical requirements on a mirror that is part of a suppression system to achieve at least 10¹⁰ rejection are extremely challenging, and a series of pathfinder demonstrations and testbeds are warranted. We examine the optical manufacturing and tolerancing requirements on the mirror itself as a function of spatial frequency where in certain regimes we desire better than 1/1000^th of a wave surface accuracy. An atypical requirement is also imposed on the optical coatings where the uniformity of reflectance is desired to be a few parts in 10,000. In addition, we present an optical design for a sub-scale coronagraphic testbed as an essential step in examining the system sensitivities.

The presence of optical aberrations in the entrance pupil of a coronagraph causes the stellar light to scatter about the occulting spot, reducing the effective contrast achievable. Even if these aberrations are sufficiently corrected with a deformable mirror to enable planet detection, small drifts in the optical alignment of the telescope introduce additional low-order aberrations. The design parameters of the coronagraph itself (e.g. occulting spot size, Lyot stop diameter, etc.) affect how these aberrations impact the contrast in the focal plane. In this study, we examine the sensitivity of contrast to low-order optical errors for several coronagraph concepts over their respective design parameters. By combining these sensitivities with the telescope throughput, we show that for each coronagraph concept there is an optimum selection of the design parameters that provides efficient, high-contrast imaging at the inner working distance in the presence of alignment errors.

The Terrestrial Planet Finder (TPF) high contrast imaging testbed (HCIT) facilitates the investigation into the diversity of engineering challenges presented by the goal of direct exo-planet detection. For instance, HCIT offers a high-density deformable mirror to control the optical wavefront errors, a configurable coronagraph to control the diffracted light, and translatable cameras for measuring the focal and pupil planes before and after the coronagraph. One of the principle challenges for a coronagraphic space telescope is the extreme level of wavefront control required to make the very faint planet signal reasonably detectable. A key component, the extremely accurate sensing of the wavefront aberrations, was recently shown to be achievable using a sufficiently constrained image-based approach. In this paper, we summarize the experimental performance a focus-diverse phase-retrieval method that uses symmetrically defocus point-spread function measurements that are obtained about the coronagraph occulter focal plane. Using the HCIT, we demonstrate the high level of wavefront sensing repeatability achieved with our particular choices of focus diversity, data fidelity and processing methodologies. We compare these results to traceable simulations to suggest a partitioning of the error sources that may be limiting the experimental results.

Motivated by the desire to image exosolar planets, recent work by us and others has shown that high-contrast imaging can be achieved using specially shaped pupil masks. To date, our masks have been symmetric with respect to a cartesian coordinate system but were not rotationally invariant, thus requiring that one take multiple images at different angles of rotation about the central point in order to obtain high-contrast in all directions. In this talk, we present two new classes of masks that have rotational symmetry and provide high-contrast in all directions with just one image. These masks provide the required 10^-10 level of contrast to within 4 λ/Δ of the central point. They are also well-suited for use on ground-based telescopes, and perhaps NGST as well, since they can accommodate central obstructions and associated support spiders.

We present the first lab experiments using a notch filter mask, a coronagraphic mask that dims the light from an on-axis source while passing off-axis light unscathed. The notch filter mask is essentially an optimized Lyot coronagraph that diffracts all of the light from the central object into a small ring that can be blocked by a Lyot stop. Notch filter masks provide a high throughput, very high contrast alternative to traditional Lyot coronagraphs. These masks, like all methods for achieving high contrast, require a high amount of accuracy in design to be successful. Nanofabrication techniques can meet these design challenges; with the first notch filter mask prototype fabricated with .25 μm precision using an e-beam lithography machine. When placed in a test bed, initial results show that 10^-5 contrast is achieved at 3λ/Δ and 10^-6 at ~8λ/Δ with a throughput of 27%. The coronagraph rejects light from the point source's peak by at least 4 orders of magnitude despite leakage of light through the mask. We speculate on the "as-is" performance of such a mask in the Hubble Space Telescope.

Recent studies of exosolar planet detection methods with a space-based visible light coronagraph have shown the feasibility of this approach. However, the telescope optical precision requirements are extremely demanding - a few Angstroms residual wavefront error - which is beyond current capabilities for large optical surfaces. Secondly, the coronagraph depends upon use of masks located at either the pupil or a focus to reject the starlight and image the exosolar planet. Effects of diffraction and light scatter place precision requirements mask manufacturing. To increase understanding and optimize performance of the coronagraph, laboratory experiments backed by end-to-end integrated models are used to project on-orbit performance. Of particular importance is the wavefront propagation through the optical system - from simple Fraunhaufer propagation to vector propagators taking into account 3D structures of the masks. Accurate models, which match test data are then used to evolve the initial coronagraph concepts for in-flight performance. In part I, we discuss error sources and model development to meet mission goals. In part II, a paper to be published at a future date, we compare lab experiment and expected residual error sources.

The TPF interferometer family suppresses the stellar glare using a deep interferometric null, which for the planet can become constructive interference because of its angular offset. The null depth need not be as great as the star-planet contrast, but its systematic fluctuations must be perhaps 5 times better than the variations which constitute the planet's signature. We present an allocation of errors which meet these needs, and identify areas which need better definition.

Polarization plays a significant role in nulling interferometry and should be considered early in the instrument design process. In this paper, I demonstrate that mirror trains with near-normal incidence angles increase the leakage into the central null more than other mirror trains for a given amount of misalignment perpendicular to the incidence plane. Also, throughput and polarization non-uniformity across a wave front will significantly increase the leakage. To simplify top-level modeling of this effect, I derive a "throughput/polarization" version of the Strehl ratio.

Data-reduction algorithms for nulling interferometers can be divided into two categories, model-fitting and imaging. We deal mostly with single-Bracewell instruments because of their simplicity, even though they suffer from “nuisance sources” such as stellar leakage and exo-zodiacal light. To simplify data reduction, we work with the Fourier compo-nents of the time series. Exo-zodiacal light dominates at low frequencies. In principle, it should be possible to model the exo-zodiacal light contribution and separate it from planets using data from a single observation. In practice, however, the uncertainty in the exact form of the exo-zodiacal cloud limits our ability to model and remove its contribution. The only unambiguous way to detect planets with a single Bracewell is to observe a system multiple times through its orbit, and look for month-to-month variations in the Fourier components. To calculate the planet parameters, we discuss a cor-relation technique based on Fourier components instead of time series, in conjunction with a linearized least-squares so-lution. Because the fringe pattern on the sky is wavelength dependent, observations over multiple bandpasses signifi-cantly increases the confidence in planet detection. These algorithms may be used with other types of nulling interfer-ometers. We briefly discuss their application to dual Bracewell data.

Deep, stable nulling of starlight requires careful control of the amplitudes and phases of the beams that are being combined. The detection of earth-like planets using the interferometer architectures currently being considered for the Terrestrial Planet Finder mission require that the E-field amplitudes are balanced at the level of ~ 0.1%, and the phases are controlled at the level of 1 mrad (corresponding to ~ 1.5 nm for a wavelength of 10 μm). These conditions must be met simultaneously at all wavelengths across the science band, and for both polarization states, imposing unrealistic tolerances on the symmetry between the optical beamtrains. We introduce the concept of a compensator that is inserted into the beamtrain, which can adaptively correct for the mismatches across the spectrum, enabling deep nulls with realistic, imperfect optics. The design presented uses a deformable mirror to adjust the amplitude and phase of each beam as an arbitrary function of wavelength and polarization. A proof-of-concept experiment will be conducted at visible / near-IR wavelengths, followed by a system operating in the Mid-IR band.

The Cold Interferometric Nulling Demonstration in Space (CINDIS) is a modest-cost technology demonstration mission, in support of interferometer architectures for Terrestrial Planet Finder (TPF). It is designed to provide as complete as possible a demonstration of the key technologies needed for a TPF interferometer at low risk, for a cost less than $300M. CINDIS foregoes scientific objectives at the outset, enabling significant cost savings that allow us to demonstrate important features of a TPF interferometer, such as high-contrast nulling interferometry at 10 μm wavelength, vibration control strategies, instrument pointing and path control, stray light control, and possibly 4-aperture compound nulling. This concept was developed in response to the NASA Extra-Solar Planets Advanced Concepts NRA (NRA-01-OSS-04); this paper presents the results of the first phase of the study.

CoRoT mission for year 2006 is a small space telescope that will measure continuously for 6 months the light flux of 12000 star in a mission of 2.5 years . The aim is to detect small droops in the light curves revealing planets transitting in front of their star. For this, 12000 logical Regions Of Interest (ROI) are defined on the CCD to optimise each star Signal to Noise Ratio (s/n). Unfortunatly only less than 256 different shapes are permitted for all ROIs, forseeing a loss in global S/N. We found a method wich reduce the 12000 ROIs to a small set of 250 shapes in a lossless way. Overall perverformances are discussed.

The COROT mission is part of the program "Petites Missions" of CNES (French space agency). It implies international cooperation between France, Belgium, Germany, Austria, Spain and the European Space Agency (ESA). COROT aims to perform astroseismology measurements and to detect exoplanets. Long duration observations of stars will be used to detect periodic variations with an afocal telescope followed by a dioptric objective and 4 CCDs. Due to the orbit of the spacecraft (low altitude polar orbit) and even if the observation are performed in a direction perpendicular to orbit plane, the measurements can be disturbed by the straylight reflected on the earth (albedo) that can generate periodic perturbation. CSL is in charge of the design and procurement, with the help of Belgian industries, of a baffle and its protective cover that will be mounted on top of the afocal entrance telescope. The requirements are very stringent from the optical point of view as well as from the mechanical point of view. The rejection of the baffle must be of the order of 10¹³ for field angles above 20 degrees while the allocated mass is 19 kilograms.

The ESA Eddington mission is designed to perform the dual science tasks of asteroseismology and exo-planet finding. The ambitious science goals can be met by a single instrument that utilises high precision and long duration photometry, that in the planet-finding phase leads the ability to detect planetary transits from targets in the magnitude range 11 - 15. We describe the baseline design of the Eddington focal plane cameras, including the performance features of e2V CCD42-C0 devices that are the proto-type detectors. Important trade-offs and system constraints are noted and issues that affect the scientific output such as radiation damage and calibration aspects, and are briefly reviewed.

Multi-object dispersed fixed-delay interferometry provides a powerful way for all sky Doppler radial velocity (RV) searches for extrasolar planets. This technique takes advantage of high sensitivity of a fixed-delay interferometer for precision Doppler RV measurements. The interferometer fringes are dispersed by a moderate resolution spectrometer for broad band observing. Compared to current state-of-the-art high resolution echelle techniques responsible for detection of more than 100 exoplanets, this technique offers several new capabilities such as wide field multiple object observation capability, high instrument throughput and stable instrument responses. The instrument can be used in the visible as well as in the near-IR and UV for observing stars with all spectral types from early A type to later M and L types. Once this kind of instrument is coupled with wide field telescopes (a few degrees of field of view, such as Sloan and WIYN), hundreds of stars with m_v < 12 or brighter can be simultaneously monitored with Doppler precision of σ = 15 m/s or better within an hour integration. Millions of stars can be monitored annually. Tens of thousands of extrasolar planets will be uncovered by this proposed all sky RV technique in a decade for studying planet formation and evolution. A prototype instrument has been observed at the KPNO 2.1m telescope in 2002, demonstrating a short term Doppler precision of ~ 3 m/s with eta Cas (V = 3.5), a RV stable star. It also helped to uncover a RV curve for 51 Peg (V = 5.5), confirming previous planet detection using the echelles. The total measured detection efficiency from the above the atmosphere to the CCD detector is about 5% without iodine absorption under 1.5 arcsec seeing conditions, comparable to all of the echelle spectrometers for planet detection. A new instrument, fed with both interferometer outputs, is being developed with a higher efficiency Volume phase holographic grating and better optical design for an initial planet search at the KPNO 2.1m telescope in 2003. It will provide about 20% total detection efficiency. In this paper, we will present principle of the technique, new results from our current instruments and grand plans for all sky Doppler surveys for extrasolar planets.

The primary limitation to ground based astronomy is the Earth's atmosphere. The atmosphere above the Antarctic plateau is different in many regards compared to the atmosphere at temperate sites. The extreme altitude, cold and low humidity offer a uniquely transparent atmosphere at many wavelengths. Studies at the South Pole have shown additionally that the turbulence properties of the night time polar atmosphere are fundamentally different to mid latitudes. Despite relatively strong ground layer turbulence, the lack of high altitude turbulence combined with low wind speeds presents favorable conditions for interferometry. The unique properties of the polar atmosphere can be exploited for Extrasolar Planet studies with differential astrometry, differential phase and nulling intereferometers. This paper combines the available data on the properties of the atmosphere at the South Pole and other Antarctic plateau sites for Extrasolar Planet science with interferometry.

This paper describes the latest progress for visible direct detection of Earth like extrasolar planets using a nulling coronagraph instrument behind a 4m class telescope. Such a system is capable of satisfying the scientific objectives of the Terrestrial Planet Finder mission In our design, a 4 beam nulling interferometer is synthesized from the telescope pupil, producing a very deep null proportional to θ⁴ which is then filtered by a coherent array of single mode fibers to suppress the residual scattered light. With diffraction limited telescope optics and similar quality components in the optical train (λ/20), suppression of the starlight to 10^-10 is achievable. Such a telescope with this nulling interferometer as back-end instrument can image and detect planets, or provide the input to a low resolution spectrometer. Shown are key features of this system in a space mission, latest results of laboratory measurements demonstrating achievable null depth, and progress toward fabrication of coherent single mode fiber arrays.

Researches have suggested several techniques (ie.: pupil masking, coronography, nulling interferometry) for high contrast imaging that permit the direct detection and characterization of extrasolar planets. Our team at JPL, in previous papers, has described an instrument that will combine the best of several of these techniques: a single aperture visible nulling corograph. The elegant simplicity of this design enables a powerful planet-imaging instrument at modest cost. The heart of this instrument is the visible light nulling interferometer for producing deep, achromatic nulls over a wide optical band pass, and a coherent array of single mode optical fibers 2 that is key to suppressing the level of scattered light. Both of these key components are currently being developed and have produced intial results. This paper will review, in detail, the design of the nulling interferometer experiment and review the latest experimental results. These results illustrate that we are well on our way to developing the fundamental components necessary for planned mission. Likewise, our results demonstrate that the current nulling levels are already consistent with final requirements.

We present the development of a single-mode spatial filter array for the nulling coronagraph application. The development consists of two generations of fiber array designs and a Zygo-interferometer based lens array to fiber array alignment methodology. We discuss the use of large mode field diameter (MFD) fibers to relax fiber placement tolerance of the fiber array. The pros and cons of using the Photonic Crystal Fiber (PCF) for building the array are discussed. The future plan for implementing a 1000-channel class, single-mode spatial filter array is described.

I describe a liquid crystal intensity modulator designed to achieve <10 parts per million (ppm) modulation to simulate a planetary transit like those required for ground testing of NASA's Kepler mission. The design uses a nematic liquid crystal as a variable retarder aligned between two linear polarizers, with the retardance values and the alignment chosen to provide low sensitivity of transmitted intensity to input liquid crystal voltage variations. Modulator test results give intensity fluctuations of a few ppm from millivolt modulations about the input 8 V baseline voltage.

An experimental proof using two liquid crystal spatial light modulators in conjunction with a white light Michelson interferometer to correct amplitude error in telescopes is presented.The principle is reviewed,and then the experiment for a monochromatic closed loop is detailed.

The Exoplanet Tracker (ET) is a new concept of instrument for measuring stellar radial velocity variations. ET is based on a dispersed fixed-delay interferometer, a combination of Michelson interferometer and medium resolution (R~6700) spectrograph which overlays interferometer fringes on a long-slit stellar spectrum. By measuring shifts in the fringes rather than the Doppler shifts in the absorption lines themselves, we are able to make accurate stellar radial velocity measurements with a high throughput and low cost instrument. The single-order operation of the instrument can also in principle allow multi-object observations. We plan eventually to conduct deep large scale surveys for extra-solar planets using this technique. We present confirmation of the planetary companion to 51Peg from our first stellar observations at the Kitt Peak 2.1m telescope, showing results consistent with previous observations. We outline the fundamentals of the instrument, and summarize our current progress in terms of accuracy and throughput.

In this manuscript, we further develop our concepts for the free-flying occulter space-based mission, the Umbral Missions Blocking Radiating Astronomical Sources (UMBRAS). Our optical simulations clearly show that an UMBRAS-like mission designed around a 4-m telescope and 10-m occulter could directly image terrestrial planets. Such a mission utilizing existing technology could be built and flown by the end of the decade. Moreover, many of the other proposed concepts for Terrestrial Planet Finder (TPF) could significantly benefit by using an external occulter. We present simultations for an optical design comprising a square aperture telescope plus square external occulter. We show that the entire diffraction pattern, which is propagated from occulter to telescope and through telescope to focal plane, may be characterized by two parameters, the Fresnel number and the ratio of the telescope diameter to the occulter width. Combining the effects of a square occulter with apodization provides a much more rapid roll-off in the PSF intensity between the diffraction spikes than may be achieved with an unapodized telecope aperture and occulter. We parameterize our results with respect to wavefront quality and compare them against other competing methods for exo-planet imaging. The combination of external occulter and apodization yields the required contrast in the region of the PSF essential for exo-planet detection. An occulter external to the telescope (i.e., in a separate spacecraft, as opposed to a classical coronagraph with internal occulter) reduces light scatter within the telescope by approximately 2 orders of magnitude. This is due to less light actually entering the telescope. Reduced scattered light significantly relaxes the constraints on the mirror surface roughness, especially in the mid-spatial frequencies critical for planet detection. This study, plus our previous investigations of engineering as well as spacecraft rendezvous and formation flying clearly indicates that the UMBRAS concept is very competitive with, or superior to, other proposed concepts for TPF missions.

Ground based adaptive optics is a potentially powerful technique for direct imaging detection of extrasolar planets. Turbulence in the Earth's atmosphere imposes some fundamental limits, but the large size of ground-based telescopes compared to spacecraft can work to mitigate this. We are carrying out a design study for a dedicated ultra-high-contrast system, the eXtreme Adaptive Optics Planet Imager (XAOPI), which could be deployed on an 8-10m telescope in 2007. With a 4096-actuator MEMS deformable mirror it should achieve Strehl >0.9 in the near-IR. Using an innovative spatially filtered wavefront sensor, the system will be optimized to control scattered light over a large radius and suppress artifacts caused by static errors. We predict that it will achieve contrast levels of 10⁷-10⁸ at angular separations of 0.2-0.8" around a large sample of stars (R<7-10), sufficient to detect Jupiter-like planets through their near-IR emission over a wide range of ages and masses. We are constructing a high-contrast AO testbed to verify key concepts of our system, and present preliminary results here, showing an RMS wavefront error of <1.3 nm with a flat mirror.

Radial velocity studies represent the most successful method to date for the detection of extrasolar planets. Although radial velocity (v_r) measurement precision of 3 m s^-1 is routinely achieved in some programs, it is important to understand and minimize sources of experimental error. Furthermore, velocity variations resulting from astrophysical processes contribute to velocity errors, and must be removed if precision is to be further improved. The use of spectrographs with telescopes having high order adaptive optics (AO) systems offers the possibility of achieving near diffraction-limited very high spectral resolution at visible wavelengths on ground-based telescopes. The small stellar image diameters obtained with adaptively corrected systems allow high resolution without a large loss of light at the spectrograph entrance aperture. The Adaptively Corrected Echelle Spectrograph (ACES), designed at Steward Observatory for a spectral resolution R ~ 200,000, couples the telescope image to the instrument with an 8-10μm diameter near single-mode optical fiber. The shorter effective slit permits the placement of more echelle orders on the detector after cross dispersion, with a correspondingly greater wavelength coverage per exposure. This simultaneous high resolution and large wavelength coverage can be used to improve the precision of radial velocity studies by improving wavelength calibration, reducing dataset internal errors, and permitting better characterization and removal of effects intrinsic to the stars themselves.

This paper describes the technical program that will demonstrate the viability of two mid-infrared nulling interferometer architectures for the Terrestrial Planet Finder (TPF) to support a mission concept downselect in 2006 between a nulling interferometer and a visible coronagraph. The TPF science objectives are to survey a statistically significant number of nearby solar-type stars for radiation from terrestrial planets, to characterize these planets and to perform spectroscopy for detection of biomarkers. A 4-telescope, 36-m Structurally-Connected Interferometer using a dual-chopped Bracewell nuller will meet the minimum science requirement to completely survey at least 30 nearby stars and partially survey 120 others. A Formation-Flying Interferometer is being designed to meet the full science requirement to completely survey at least 150 stars, and involves a trade between dual-chopped Bracewell, degenerate Angel Cross, and the Darwin bow-tie configuration. The system engineering trades for the connected structure and formation-flying architectures are described. The top technical concerns for these architectures are mapped to technology developments that will retire these concerns prior to the project downselect.

Nulling interferometry shows promise as a technique enabling investigation of faint objects such as planets and exo-zodiacal dust around nearby stars. At Jet Propulsion Laboratory, a nulling beam combiner has been built for the Terrestrial Planet Finder project and has been used to pursue deep and stable narrowband nulls. We describe the design and layout of the modified Mach Zehnder TPF nuller, and the results achieved in the laboratory to date. We report stabilized nulls at about the 10^-6 level achieved using a CO₂ laser operating at 10.6 μm, and discuss the alignment steps needed to produce good performance. A pair of similar nullers has been built for the Keck Observatory, for planned observations of exo-zodiacal dust clouds. We also show briefly a result from the Keck breadboard experiments: passively stabilized nulls centered around 10.6 micron of about 2 10^-4 have been achieved at bandwidths of 29%.

Detection of extra-solar, and especially terrestrial-like planets, using coronagraphy requires an extremely high level of wavefront correction. For example, the study of Woodruff et al. (2002) has shown that phase uniformity of order 10^-4λ(rms) must be achieved over the critical range of spatial frequencies to produce the ~10¹⁰ contrast needed for the Terrestrial Planet Finder (TPF) mission. Correction of wavefront phase errors to this level may be accomplished by using a very high precision deformable mirror (DM). However, not only phase but also amplitude uniformity of the same scale (~10^-4) and over the same spatial frequency range must be simultaneously obtained to remove all residual speckle in the image plane. We present a design for producing simultaneous wavefront phase and amplitude uniformity to high levels from an input wavefront of lower quality. The design uses a dual Michelson interferometer arrangement incorporating two DM and a single, fixed mirror (all at pupils) and two beamsplitters: one with unequal (asymmetric) beam splitting and one with symmetric beam splitting. This design allows high precision correction of both phase and amplitude using DM with relatively coarse steps and permits a simple correction algorithm.