Complex non-linear and dynamic processes lie at the heart of the planet formation process. Through numerical simulation and basic observational constraints, the basics of planet formation are now coming into focus. High resolution imaging at a range of wavelengths will give us a glimpse into the past of our own solar system and enable a robust theoretical framework for predicting planetary system architectures around a range of stars surrounded by disks with a diversity of initial conditions. Only long-baseline interferometry can provide the needed angular resolution and wavelength coverage to reach these goals and from here we launch our planning efforts. The aim of the Planet Formation Imager" (PFI) project is to develop the roadmap for the construction of a new near-/mid-infrared interferometric facility that will be optimized to unmask all the major stages of planet formation, from initial dust coagulation, gap formation, evolution of transition disks, mass accretion onto planetary embryos, and eventual disk dispersal. PFI will be able to detect the emission of the cooling, newlyformed planets themselves over the first 100 Myrs, opening up both spectral investigations and also providing a vibrant look into the early dynamical histories of planetary architectures. Here we introduce the Planet Formation Imager (PFI) Project (www.planetformationimager.org) and give initial thoughts on possible facility architectures and technical advances that will be needed to meet the challenging top-level science requirements.
We have built at the Haute-Provence observatory (France) the rst diluted telescope in the world. We describe
this prototype called Carlina, made of three 25 cm mirrors separated by a maximum baseline of 10.5 m. The
three mirrors in place are already coherenced and rst light is scheduled for June-July 2012. In this article, we
will mainly describe the focal gondola. We propose to build in the near future a 100 m aperture Large Diluted
Telescope. This diluted telescope will be more sensitive than regular interferometers (Keck, VLTI, etc.), with
higher imaging capabilities. A LDT will open new elds of research in astrophysics thanks to very high angular
resolution imaging of the surface of supergiant stars, AGN, gravitational micro-lens systems, exo-planets, etc.
In a previous paper,1 we discussed an original solution to improve the performances of coronagraphs by adding,
in the optical scheme, an adaptive hologram removing most of the residual speckle starlight.
In our simulations, the detection limit in the flux ratio between a host star and a very near planet (5λ/D)
improves over a factor 1000 (resp. 10000) when equipped with a hologram for cases of wavefront bumpiness
imperfections of λ/20 (resp. λ/100).
We derive, in this paper, the transmission accuracy required on the hologram pixels to achieve such goals. We
show that preliminary tests could be performed on the basis of existing technologies.
For imaging faint and complex sources at high angular resolution, hypertelescopes (direct-imaging many-aperture
interferometers using a densified pupil) gain sensitivity with respect to few-aperture interferometers and to Fizeau
interferometers. Steps are taken to expand the Carlina-Proto technical prototype built at Observatoire de Haute-Provence,
18m in aperture size, and to define a larger (100-200m) Carlina-Science version, incorporating 100 or more small
apertures. Following initial observing by Speckle Interferometry, adaptive co-pistoning is expected to become available,
using "Dispersed Speckle" piston sensing on bright stars, and a modified Laser Guide Star on faint (mv > 25) fields.
"Extremely Large Hypertelescope" versions of such instruments, with aperture size beyond a kilometer, are considered
for deep-field imaging on cosmological sources. These can be interferometrically coupled with ELTs, or arrays of
telescopes, at sites such as the Macon range (Andes) considered by ESO for its E-ELT. Space versions are proposed to
ESA and NASA.
SOPHIE is a new fiber-fed echelle spectrograph in operation since October 2006 at the 1.93-m telescope of Observatoire
de Haute-Provence. Benefiting from experience acquired on HARPS (3.6-m ESO), SOPHIE was designed to obtain
accurate radial velocities (~3 m/s over several months) with much higher optical throughput than ELODIE (by a factor of
10). These enhanced capabilities have actually been achieved and have proved invaluable in asteroseismology and
exoplanetology. We present here the optical concept, a double-pass Schmidt echelle spectrograph associated with a high
efficiency coupling fiber system, and including simultaneous wavelength calibration. Stability of the projected spectrum
has been obtained by the encapsulation of the dispersive components in a constant pressure tank. The main
characteristics of the instrument are described. We also give some technical details used in reaching this high level of
The coronagraphic techniques serving to reject most light from a star, when trying to image a nearby planet, can be pushed with an adaptive holographic element. Located after the coronagraph, it can in principle remove most of the residual star light by adding a phase-shifted holographic reconstruction of it . The scheme is also usable within each sub-aperture of a diluted hypertelescope array, sufficiently large to resolve details of an exo-Earth. A possible panoramic version of the previously mentioned Exo-Earth Imager is shaped as a virtual bubble of 400 km diameter , consisting of thousands of 3-meter mirrors, free-flying and arranged co-spherically. The half-size focal sphere is explored by beam combiners, one for each exo-Earth observed within tens of parsecs. Each beam-combiner includes a kilometer-sized corrector of spherical aberration at F/2, which is also diluted and consisting of small free-flyers. The instrument is expected to provide direct coronagraphic images of exo-Earths, resolved in 50x50 resels, with enough dynamic range obtained in 30mn exposures to search colored features and their seasonal variations, indicative of photosynthetic life .
'Densified-pupil multi-aperture imaging arrays', also called hypertelescopes, provide a path towards rich images obtained directly at the focal plane. They typically involve a large Fizeau arrangement, with a small attached 'pupil densifier' serving to gain luminosity at the expense of field. At scales ranging from kilometers to perhaps a million kilometers, such architectures appear of interest for stellar physics, galaxies, cosmology, and neutron star imaging with the larger sizes. Ground testing is initiated and space versions are proposed, particularly to NASA for its Terrestrial Planet Finder. The coronagraphic imaging achievable with this space version is expected to improve the detection sensitivity to attenuating the sky background contribution. Subsequent laser versions can in principle resolve the 'green spots' on an Earth seen at several parsecs. Current design work for a precursor array of 'flying mirrors' driven by solar sails in geostationary orbit will be presented.