
In order to make such observations, one of the key tools will be an optical telescope of unprecedented size. The largest telescope planned is the European Extremely Large Telescope, the primary mirror of which will be 42 metres in diameter. Such a telescope must of necessity be ground-based rather than space-based and yet its imaging qualities must approach those of a space telescope. This means overcoming the effects of looking through the inevitable atmospheric turbulence above the telescope. The technology to do this has been under the development for the last twenty years or so and is called Adaptive Optics. This technique uses special deformable mirrors that can adapt themselves to cancel the evolving atmospheric distortions of the light wave-fronts at frequencies of many hundreds of times per second. Hitherto, the Adaptive Optics technique has only operated over fairly limited fields of view. To survey the early Universe efficiently, however, larger areas of sky must be corrected all at once. This is the role of the EAGLE instrument, which is being designed in France and the United Kingdom (UK), and will be able to look in detail at 20 targets at a time, each one of which will be separately corrected for atmospheric distortion. However, the first galaxies are too faint to allow us to measure the atmospheric distortions. It is therefore required to find in the large field of view a number of guide stars bright enough for this purpose. It is also envisioned to generate artificial laser guide stars when the number of available natural guide stars is not sufficient. Moreover, using, as in conventional adaptive optics, only one deformable mirror for compensating for the whole field of view is not feasible. Because of the very small size of the first galaxies (of the order of one arc second), it is instead only required to compensate for the turbulence in a few small field islands (the target galaxies) distributed throughout the whole field of view. The new technology, called Multi-Object Adaptive Optics (MOAO), implements one deformable mirror per target in a dedicated optical train, all the deformable mirrors being controlled in open loop (no feedback), from all the in-the-field measurements performed on all the guide stars, through a global tomography approach to measuring the atmosphere. This new technique has just been demonstrated on-sky for the first time by members of the EAGLE team at the William Herschel Telescope on La Palma in the Canary Islands, Spain. The William Herschel Telescope has a 4.2 metre diameter primary mirror and is therefore exactly one tenth of the size of the future European Extremely Large Telescope. It is, however, still rather large by current standards and has particular features which make it very attractive for evaluating new technologies. In particular it has two massive platforms were experimental systems can be mounted in a stable and easily accessed environment. The new technical demonstrator system for EAGLE is called CANARY, reflecting both its relative size and the fact that it is used on the Canary Islands. CANARY was used for the first time on the William Herschel Telescope and the results were a spectacular success. The new technology of Multi-Object Adaptive Optics worked the first time it was used, and furthermore it operated at the hoped-for performance level. Figure 1 shows the image of a test star with the correction system switched off and then with it working. The achieved performance with MOAO is very similar to the one obtained with the conventional adaptive optics mode. The corrections were applied 150 times every second using information gathered from three other off-axis stars for the wavefront measurements some distance from the on-axis test star. This information was used to infer the correction required in the direction of the test star where the deformable mirror is installed. It is this crucial tomography of the atmosphere above the telescope and the deformable mirror control in open loop that are the novel achievements of CANARY. The next step will be to repeat this experiment but using artificial laser guide stars generated in the atmosphere using powerful lasers. This technique has been used elsewhere, but never to perform the tomographic measurements required by EAGLE and hence CANARY. The use of lasers stars is essential for EAGLE as there are not sufficient real stars of the required brightness to allow the full survey to be done.

See the page of INSU-CNRS The institutions involved in the development and the first demonstration of CANARY are : - Laboratoire d’Etudes Spatiales et d’Instrumentations en Astrophysique (LESIA) (Observatoire de Paris, CNRS, UPMC et Université Paris Diderot), France - Laboratoire d’Etudes des Galaxies, Etoiles, Physique et Instrumentation (GEPI) (Observatoire de Paris, CNRS et Université Paris Diderot), France - Durham University, UK - Astronomy Technology Centre (UKATC), UK Professor Gerard Rousset of LESIA and Doctor Richard Myers of Durham University are the co-leaders of the CANARY Project. The CANARY team also acknowledges the contributions from Engineering and Project Solutions Ltd and from staff currently at PUC, Chile. There has also been a completely invaluable contribution from the staff of the Isaac Newton Group of telescopes, who operates the William Herschel Telescope, which made this demonstration possible. The institutions involved in EAGLE are : - Laboratoire d’Astrophysique de Marseille (LAM) (CNRS, Université de Provence), France - LESIA - GEPI - Office National d’Etudes et de Recherches Aéropsatiales (ONERA), France - UKATC - Durham University EAGLE is led by Doctor Jean-Gabriel Cuby of LAM and Professor Simon Morris of Durham University. In the subsequent (laser) phases, all the EAGLE members will also contribute to CANARY. We are grateful for funding from : - Agence Nationale de la Recherche (ANR) programme 06-BLAN-0191, - Centre National de la Recherche Scientifique (CNRS) - Institut National des Sciences de l’Univers (INSU), - Observatoire de Paris, - et Université Paris Diderot - Paris 7, en France ; - Science and Technology Facilities Council - et Durham University au Royaume-Uni ; - Commission Européenne (Framework Programme 7) avec : E-ELT Preparation, Infrastructures 2007-1 Grant 211257 et OPTICON Infrastructures 2008-1 Grant 226604. Contact Gerard Rousset (Observatoire de Paris, LESIA et CNRS) Eric Gendron (Observatoire de Paris, LESIA et CNRS)