Research Network for Adaptive Optics
Current Status of International Research
|
Australia contributes to a global effort to research and develop
technologies for implementing adaptive
optics systems and for using them in diverse applications. Here we
address the current state of international research into
adaptive optics technologies and its end-uses, and how research
activity in these areas are likely to develop over the coming
decade.
Optics research is experiencing a resurgence with the advent of adaptive correction to optical systems. Adaptive optics systems have become an integral part of the current generation of large optical/infrared telescopes. They can now produce diffraction-limited images at near-infrared wavelengths that are as sharp as the visible wavelength images obtained by the Hubble Space Telescope. The first Multi-Conjugate Adaptive Optics (MCAO) systems are under development. These systems use multiple deformable mirrors, each targetting layers of turbulence at different heights in the Earth's atmosphere, to achieve good imaging performance over a wide field of view (approximately one arcminute of the sky or 1/30th of the diameter of the Moon). Adaptive optics systems have been incorportated into laboratory scanning laser ophthalmoscopes to image living retinal tissue for the first time at resolutions sufficient to resolve individual cones. Spectacular results have been obtained showing that the relative numbers of red, green, and blue sensors differ among different people. Free space optical communication, joining optical fibre communications networks via secure, portable, high-bandwidth laser telemetry, is a reality. Adaptive optics is being explored to increase communications range, overcome atmospheric turbulence, and track building sway. Laser telemetry has already been demonstrated for secure communications between spacecraft. Laser countermeasures are being actively developed for military applications. The Research Network for Adaptive Optics embraces the following research areas:
These are all classical adaptive optics systems that use a natural guide star to sense wavefront distortion and a single deformable mirror to correct that distortion. These systems produce diffraction-limited images over only a small field of approximately 20 arcseconds radius around a bright star, and so they are limited in the fraction of sky over which they can be used. The image shape, or point spread function, varies over the field so while they are capable of obtaining spectacular images, calibration of absolute photometry is difficult. Astronomical adaptive optics is still a young field. Many of the systems are experimental, so systematic effects like residual image artifacts are problematic and have led to mis-interpretation of scientific results. All of these systems require expert users to achieve optimal results.
 Intensity distribution in a star image; uncorrected (left), AO-corrected (right). Courtesy Uni. of Hawaii Laser guide stars offer a way of overcoming the sky coverage limitation of natural guide stars. There are two sorts; Rayleigh beacons and sodium laser guide stars. Rayleigh beacons are formed by backscatter from a laser beam in the lower atmosphere below altitudes of about 10 km. These require lower power and were developed early on but have been resurrected recently by UK groups aiming to correct only turbulence in the lower atmosphere. Rayleigh beacons are limited in their application because the cone illuminated by these low-altitude laser guide stars does not sample all of the atmosphere traversed by the wave fronts from an astronomical object. Sodium laser guide stars are formed by exciting atmospheric sodium atoms in a layer approximately 10 km thick at an altitude of about 90 km. Sodium laser guide stars require powerful lasers (about 10W) locked to the Na I D line frequency. It has proven to be extremely difficult to develop robust, reliable lasers that fulfil these requirements. Sodium dye lasers have been operated at Lick Observatory in California, Calar Alto in Spain, and at the Keck Observatory in Hawaii. However, these lasers are very labour-intensive to maintain. There is currently a world-wide effort underway to develop cost-effective, reliable sodium lasers for this application.
 Rayleigh beacon launched from the William Herschel Telescope. Courtesy Uni. of Durham AO group The technique of multi-conjugate adaptive optics is under development to overcome the other limitation of classical astronomical adaptive optics systems, i.e., limited diffraction-limited field. In essence, the degree of wavefront correction degrades away from a guide star because atmospheric turbulence becomes successively less correlated the further one moves away from the guide star in angle on the sky. In principle, turbulence close the the telescope produces similar distortions across the telescope aperture no matter which direction one looks, but at higher altitudes the light beams from separated objects pass through different air volumes. Classical adaptive optics systems with a single deformable mirrors correct wave front errors only on axis. Multi-conjugate adaptive optics systems use several deformable mirrors and wave front sensors optically matched to different altitudes to build a three-dimensional picture of the wave front distortions and correct them. In principle, this atmospheric tomography can achieve similar image quality (i.e., similar point spread functions) over angular fields of approximately 1 arcminute. Multi-conjugate adaptive optics has been demonstrated in the laboratory, but no system has yet been demonstrated on a telescope. However, several systems are currently under development.
 Multi-conjugate adaptive optics principle. Courtesy ESO AO group. The Gemini Observatory is developing a Multi-Conjugate Adaptive Optics system for the 8-m diameter Gemini South telescope. This system will use five laser guide stars and three natural guide stars. The Research School of Astronomy and Astrophysics of The Australian National University's is building the science casmera for this wide-field adaptive optics system, and Electro Optics Systems Ltd in Queanbeyan NSW are building the AO Module through their EOS Technologies Inc. subsidiary in Tucson USA. The European Southern Observatory is developing a Multi-Conjugate Adaptive Optics Demonstrator (MAD) consisting of two deformable mirrors to test MCAO principles and wave front sensor technmologies. These multi-conjugate adaptive optics systems have taken on greater importance since the recent decision by NASA to cancel the planned servicing mission for the Hubble Space Telescope. Multi-conjugate adaptive optics systems on ground-based large telescopes are now the only long-term prospect for obtaining high spatial resolution images to support future facilities such as the Jame Webb Space Telescope that is set to be launched towards the end of this decade. This will operate only at infrared wavelengths. Even larger optical/infrared astronomical telescopes with diameters between 30-m and 100-m are being planned in the USA, Canada, and Europe. These facilities are expected to be available around 2015. While the details of these developments are beyond the scope of this review, what is clear is that all of these Extremely Large Telescopes (ELTs) will require adaptive optics systems to realise the extremely high diffraction-limited spatial resolutions that are possible with telescopes of this aperture, and which are essential if these facilities are to realise their science goals. The development of these adaptive optics systems is a major undertaking in all countries aspiring to build ELTs. It is not know whether current adaptive optics principles scale to the larger apertures of ELTs. Certainly, Kolmogorov power-law turbulence profiles break down beyond the outer turbulence scale length, which some measurements suggest may be around 20 m. It is also by no means clear that optimal techniques for adaptive optics on 8-m diameter telescopes are necessarily the best approaches for a 100-m diameter.
 ESO's 100-m diameter OverWhelmingly Large (OWL) Telescope. Courtesy ESO. Four variants of adaptive optics systems are proposed for the next generation of ELTs; classical single-conjugate adaptive optics (SCAO) systems, multi-conjugate adaptive optics (MCAO) systems, ground-layer adaptive optics (GLAO) systems, and extreme adaptive optics (ExAO) systems. SCAO and MCAO systems will fulfill their current functions but require many more deformable mirrors actuators across the larger telescope apertures. GLAO systems aim to achieve only partial correction over fields of several arcminutes. These systems offer improvements in image diameter of a factor of 2-3 over natural seeing over the full telescope field. They are aimed at improved performance for wide field spectroscopic surveys. A major difficulty is distinguishing between the effects of ground-layer turbulence, which is the same for all field positions and must be corrected, and high altitude turbulence, which differs for different field positions and degrades the field-averaged performance if corrected. Low-altitude Rayleigh beacon laser guide stars have been proposed as a means of sampling only the low-altitude turbulence. ExAO systems, on the other hand, are required to directly image planets around nearby bright stars. Contrast ratios in excess of 107 are required to do this so ExAO systems must achieve high-order correction on axis using many thousands of deformable mirror actuators. These systems are extremely complex in terms of the number of deformable mirrors actuators, the number of wavefront sensor channels, and the complexity of the computer system and algorithms needed to close the adaptive optics control loop at speeds in excess of 1 kHz. These systems are the pinnacle of current adaptive optics development internationally. Both ExAO and GLAO systems are among the second generation instruments for the Gemini telescopes that are currently undergoing design studies. These systems are less complex than those needed for the larger ELTs, but will serve to define many of the approaches that will be required to develop the ELT systems. As a result of pioneering Australian research, it is now recognised that Antarctic plateau sites have atmospheric turbulence properties that are fundamentally different to those at temperate locations. This has led several research groups, including those at the University of Arizona and the Osservatorio Astrofisico di Arcetri, to perform model calculations of the performance of a hypothetical ELT located in the Australian Antarctic Territory at Dome C. Research is continuing to assess whether the likely gains in performance and/or simplification of the adaptive optics system are sufficient to compensate for the remoteness of the Antarctic location. Current astronomical adaptive optics systems sense wave front shape using the light of a few individual guide stars. These star-oriented wave front sensors are inefficient in the sense that all objects in the field contain information about the wave front shape. A new approach has been proposed call layer-oriented wave front sensing. In essence, the wave front is sensed at an image of the telescope primary mirror (or close to it) rather at the focus, so that the light from all objects in the field is combined. Different layer-oriented wave front sensors can be made to sample the wave front at different altitudes, and so command multiple deformable mirrors more directly. This approach may avoid the need for laser guide stars on Extremely Large Telescopes.
 Star-oriented (left) and layer-oriented (right) wave front sensing. Courtesy Oss. Ast. di Arcetri The Gordon and Betty Moore Foundation has recently donated $US9.1M to the Center for Adaptive Optics at the University of California at Santa Cruz to establish the Moore Laboratory for Adaptive Optics. The Laboratory for Adaptive Optics should be operational by February 2004. It will develop innovative instrumentation for the application of adaptive optics technology in astronomy. This will include simulation of the conditions that cause point spread function degradation in ExAO and simulation of the performance of multi-conjugate adaptive optics.
 Image of the retina without (left) and with (right) adaptive optics. Image by Austin Roorda. The scientific potential of these instruments is still being defined. Early studies have revealed new information about the organisation and function of the retina and the optical and neural limits of human vision. The arrangement and relative numbers of the three cone classes in the retina have been studied. These are the photoreceptors responsible for colour vision; long wavelength (L), medium wavelength (M), and short wavelength (S) photoreceptors. The existence of L-, S-, and M-photoreceptors is confirmed and they have been shown to be randomly distributed in the retina with the relative numbers of L- and M-cones differing greatly between subjects have normal and similar color vision. The resolution of the eye is potentially affected by retinal resolution and neural factors as well as by optical aberrations. Studies with adaptive-optics corrected instruments suggest that even sharper resolution can be obtained by correcting higher-order aberrations in the eye. Spectacles have been used for centuries to correct defocus. It is only in the last 150 years that spectacles have been used to also correct astigmatism. Adaptive optics systems offer the possibility of accurately measuring high-order aberrations in individuals and customizing the correction to each person. Methods for implementing this are currently being explored; customized contact lenses, intraocular lenses, or customized laser refractive surgery. The ability to images individual photoreceptors in the living eye may also improve our ability to diagnose eye disease. Glaucoma, a disease in which blindness ensues from the gradual loss of optic nerve fibres, can only be detected with conventional imaging techniques after a significant amount of damage has already occurred. More accurate measurement of the nerve fibre layer thickness around the optic nerve head will allow earlier detection of this disease. Diabetes is another disease that affects the retina by the formation of microaneurysms in the retinal vasculature. Treatement requires accurate delivery of a photocoagulating laser beam, which could benefit from a higher resolution view of the critical retinal region. The original adaptive optics corrected ophthalmic instruments used low-order deformable mirrors for aberration correction; typically 37-actuator faceplate mirrors. Higher order correction systems are now in use. These typically use 97-actuator deformable face-plate mirrors that are so large that they require elaborate beam expansion optics. These systems are expensive and bulky, which makes them unsuitable for clinical applications. The push in this area is now towards Micro Electro-Mechanical Systems (MEMS) deformable mirrors that are much smaller and cheaper than piezo-electrically driven face-plate mirrors. The availability of suitable MEMS devices will revolutionise the market for clinical adaptive optics corrected scanning laser ophthalmoscopes. At the moment, the main limitation of MEMS deformable mirrors is their stroke (i.e., the amplitude of wavefront error that can be corrected). This is currently only a few microns, which is smaller than desirable.
 Laboratory confocal scanning laser ophthalmoscope. Courtesy William's Lab.
Several groups are developing commercial adaptive optics corrected free-space optical communications products; e.g., AOptix Technologies Inc, and the ALFONSO consortium in the UK. With the downturn in the global telecommunications industry, this effort is being redirected towards secure communications for military applications. Laser telemetry is used for spacecraft communication where it benefits for adaptive optics correction to reduce laser beam divergence and provide dynamic beam tracking.
It is often desirable to dynamically control low-order aberrations in laser systems. One example is focus control to adapt to height variations in a micro-machining sample. Systems for doing this have been tested overseas, but are not yet commonly available.
 Airborne Laser in operation. Raytheon and Lockheed-Martin are developing a directed energy weapon for the F-35 Joint Strike Fighter. The system will use a 100 kW solid state laser and adaptive optics to correct for near-aircraft turbulence that would otherwise distort the laser beam.t is expected to have a range of between 6.5 km and 10 km.
The focus in the astronomy community is on higher-order deformable mirrors with larger numbers of actuators to deliver better correction of atmospheric turbulence. MEMS deformable mirrors are of interest because of their lower cost and the potential for reducing the overall size of the adaptive optics system. However, the push is also to wider fields which pushes back towards larger systems. Adaptive optics systems have typically been placed at the telescope focus so require additional optics to reform the pupil image. Their are significant advantages to integrating the deformable mirrors into the telescope itselfs, albeit with added complexity since these mirrors at typically aspheric. A 336-actuator adaptive secondary mirror has been built by the Osservatorio Astrofisico di Arcetri, Italy, and Steward Observatory, USA, and is now operational on the 6-m diameter MMT in Arizona. The challenge now is to manufacture large deformable mirror segments that are cost-competitive with rigid mirrors. More efficient wave front sensors are sought. The Osservatorio Astrofisico di Arcetri adaptive optics group has developed a Pyramid Wave Front Sensor that offers some advantages over Shack-Hartmann wave front sensors. These are under test. As the number of actuators increases, so too does the complexity of the real-time reconstructor. These is especially the case for Multi-Conjugate Adaptive Optics systems. Novel sparse matrix inversion techniques are being developed and implemented to handle this. Astronomical adaptive optics systems will benefit from laser guide stars. There is a global effort to perfect robust sodium lasers for this purpose. Sodium laser guide stars have the undesirable characteristic of appearing elongated with viewed by subapertures that are significantly off axis. This is because the atmospheric sodium layer has a thickness of approximately 10 km. This effect will be especially problematic for the next generation of Extremely Large Telescopes. It can be overcome using a pulsed laser and tracking the light pulse as it passes through the sodium layer. The Association of Universities for Research in Astronomy in the USA has recently funded a group led by the Keck Observatory to develop the next generation of optical detectors for wavefront sensors. These detectors will have an architecture that permits this laser pulse tracking. The focus in ophthalmic adaptive optics is much more towards compact systems using MEMS deformable mirrors. Higher order correction is desirable for research applications, but the major impact will be in clinical instruments. Other adaptive optics applications are either in early phases of development or use techniques developed principally for astronomic and ophthalmic applications. |