Progress in adaptive optics

Astronomers need to survey up to 40 objects simultaneously with improved resolution and select them over a wide field of view. This ‘multiplex gain’ is required for key surveys of the first stars and to determine the assembly mechanism of galaxies. It can also be used for detailed studies of more recent stellar populations in galaxies external to our own.

Adaptive optics (AO) technology is used in some form on most 8-meter class optical telescopes and on some smaller telescopes. It is a method for reducing the blurring effects of the atmospheric turbulence above such telescopes. It detects atmospheric aberrations using either natural or artificial ‘laser guide stars’ (LGSs: see Figure 1) that are created artificially in the atmosphere by projecting a beam from the telescope. The backscattered light serves to illuminate and measure the atmospheric turbulence, and it can be used where no adequately bright natural guide star is available near the scientific target. Deformable mirrors that can be reshaped several hundred times per second are then used to correct the signal. AO is particularly effective at near-IR wavelengths because the stringency of the required wavefront phase correction is considerably less than for shorter (visible) wavelengths. At still longer wavelengths, adaptive correction becomes less necessary because the phase distortions from the atmosphere are reduced (although the actual path length changes are the same).

Figure 1. The William Herschel Telescope Rayleigh Laser Beacon. Photo: Tibor Agocs, Isaac Newton Group of Telescopes.

First generation AO systems correct only a limited field of view around the guide star, but the needs of astronomers have driven the technology towards increasing the corrected field. Multi-conjugate adaptive optics^1,2 can produce a uniformly high order of correction over a wide field of view: around two arcminutes diameter, rather than up to a few tens of arcseconds for a first-generation AO system. (There are 60 arcseconds in an arcminute and 60 arcminutes in one degree.) If an even larger corrected field is required, as it is for some key astronomical observations, then ground layer AO (GLAO) and multi-object AO (MOAO) can provide solutions. We are developing these new technologies for existing 4-meter and 8-meter telescopes. They are of interest for future large telescopes of up to 42-meter diameter. The corrected fields in this case can be as much as ten arcminutes diameter.

GLAO relies on the observation that the strength of turbulence is not uniformly distributed vertically. On the contrary, the strongest turbulence tends to be concentrated in the first kilometer above the telescope. If this ground layer of turbulence could be selectively corrected with a deformable mirror, then the partial correction would be applicable over a wide field of view. The correction would be partial because only part of the turbulent volume would be corrected. The key to the technique is to selectively sense the ground layer turbulence. On 4m telescopes this can be achieved using a single low-altitude ‘Rayleigh’ laser guide star. This form of GLAO has been implemented on the William Herschel and Southern Astrophysical Research (SOAR) telescopes, where the GLAO systems are known as GLAS³ and SAM,⁴ respectively. My team contributed system modeling, laser diagnostics, real-time control and commissioning to the GLAS GLAO system.

For 8m telescopes, the sensing is achieved by multiple LGSs. Here the signals from the multiple LGSs are compared and the common ground layer signal extracted. In the case of the European Organisation for Astronomical Research in the Southern Hemisphere (ESO) Adaptive Optics Facility⁵ for its Very Large Telescope (VLT), this will be achieved with four high-altitude ‘sodium’ LGSs. In the case of the Large Binocular Telescope ARGOS⁶ AO system, three Rayleigh laser guide stars will be used on each of the two eyes of the telescope. Both systems are under design and construction and both use adaptive secondary mirrors to implement the correction. GLAO is also a planned AO mode for the 42-meter European Extremely Large Telescope (E-ELT).⁷

What if the astronomical observation requires higher order correction and an extended (10-arcminute) field of view? This is indeed the case for certain key cosmological and other survey investigations, and MOAO is a new concept that aims to meet this requirement. The sensing arrangement superficially resembles the multi-LGS GLAO case, but the whole turbulent volume above the telescope is determined, rather than just the ground layer. Individual lines of sight to astronomical targets are projected through this volume by the control system, and the resulting bespoke corrections fed to separate deformable mirrors for each scientific target. Unusually, the correction for these mirrors must be applied in an open-loop fashion (i.e., with the correction not seen by the sensing system).

This form of AO has been studied for 8m telescopes in the FALCON⁸ study and for the thirty-meter telescope (TMT).⁹ FALCON was an MOAO design study for one of the 8m telescopes of the ESO VLT. It is also key to the EAGLE¹⁰ MOAO-enabled large survey field spectroscopic instrument being studied for the E-ELT. My team contributed to the EAGLE Phase A study. We worked on the design of the laser guide star wavefront sensors, end-to-end modeling, and the design of real-time control hardware.

Because of the novelty of MOAO in terms of open-loop control, high-accuracy tomography and calibration, several laboratory and on-sky demonstrator projects are underway, such as SESAME (the Paris Observatory MOAO test bench),¹¹ the UCO/Lick MCAO/MOAO testbed,¹² Villages (a UCO/Lick on-sky open-loop AO demonstrator),¹³ VOLT (a University of Victoria Open-loop testbed),¹⁴ and RAVEN (their new on-sky open-loop MOAO system).

The CANARY project, which I lead, is specifically an on-sky demonstrator for the EAGLE E-ELT instrument.¹⁵ It will investigate the LGS tomography and calibration problems by building a 1:10 scale demonstrator of an EAGLE MOAO channel on the 4.2-meter William Herschel Telescope in Spain. The first, NGS only, variant of CANARY will go on-sky in July of 2010, with Rayleigh LGS variants following in 2011 and 2012.

The next steps will include the validation of proposed technical improvements, such as combining laser-guide star information from different altitudes to enhance correction at shorter wavelengths. There will also be scientifically capable pathfinder instruments which will do the first MOAO astronomy. One possibility is CONDOR, the 8m MOAO instrument proposed in Chile.

CANARY is a UK-France project with participation from the Paris Observatory, the French Aerospace Lab, the Astrophysics Laboratory of Marseille, the United Kingdom Astronomy Technology Centre, Durham University, and the Isaac Newton Group of Telescopes on the Canary Islands, Spain.

The CANARY project is grateful for the support from UK and French funding agencies, including the Science and Technology Facilities Council in the UK, the National Center for Scientific Research and the National Agency for Research (grant 06-BLAN-019) in France, and from the European Union through the 7th Framework Programme Extremely Large Telescope Preparation and Optical Infrared Coordination Network for Astronomy funding mechanisms.