An introduction to Active & Adaptive Optics
Ground-based Optical/Near-Infrared large Telescopes are crucial tools for the understanding of our Universe, but their image quality is severely limited by the (quasi-static) errors in the telescope itself and the (very dynamic) atmospheric turbulence inside and over the telescope. Active Optics is used to overcome the first limitation and Adaptive Optics the latter, giving ultimately images near the diffraction limit of the primary mirror. There are a number of physical limitations to adaptive optics performance, leading to successive generations of more and more sophisticated techniques detailed below.
Since its invention 400 years ago, the astronomical telescope has evolved from a small, manually pointed device for visual observations to a large and sophisticated computer-controlled instrument with full digital output. Throughout this development, two parameters have been particularly important: the light-collecting power or diameter of the telescope (allowing the detection of fainter and more distant objects) and the angular resolution (or image sharpness). For a perfect telescope used in a vacuum, resolution is directly proportional to the inverse of the telescope diameter. A plane wavefront from distant star (effectively at infinity) would be converted by the telescope into a perfectly spherical wavefront, forming the image, with an angular resolution only limited by light diffraction - aptly called the diffraction limit.
In practice, however, both atmospheric and telescope errors (Fig.1) distort the spherical wavefront, creating phase errors in the image-forming ray paths. Even at the best sites, ground-based telescopes observing at visible wavelengths cannot, because of atmospheric turbulence alone, achieve an angular resolution better than telescopes of 10- to 20-cm diameter. For a 4-m telescope, atmospheric distortion degrades the spatial resolution by more than one order of magnitude compared with the diffraction limit, and the intensity at the center of the star image is lowered by a factor of 100 or more. The cause is random spatial and temporal wavefront perturbations induced by turbulence in various layers of the atmosphere; one of the principal reasons for flying the Hubble Space Telescope was to avoid this image smearing. In addition, image quality is affected by permanent manufacturing errors and by long time scale-wavefront aberrations introduced by mechanical, thermal, and optical effects in the telescope, such as defocusing, decentering, or mirror deformations generated by their supporting devices.
Until recently, the astronomical telescope has remained a "passive" instrument. Without any in-built corrective devices to improve the quality of star images during observations, the only possible adjustments were those performed during daytime or at the beginning of the night.
Although it was thought that atmospheric distortions could not be avoided, mechanical improvements have been made to minimize telescope errors. Mirror figuring and polishing were improved, and stiffer structures and mirrors used to minimize gravitationally-induced deformations. Low-expansion glass was introduced to avoid mirror distortions as temperature varies. To reduce local temperature effects, heat dissipation from motors and electronic equipment was minimized during the night, and the dome, which in addition shields the telescope from the effects of wind buffeting, cooled during the day. In such properly designed and well-manufactured medium size telescopes, image quality is limited mainly by atmospheric distortions.
However, as plans were developed in the 80s to enhance light-collecting power by building telescopes with primary mirrors well above 4 m in diameter, it became clear that conventional methods of maintaining image quality were ruled out by cost and structure weight limitations. As a result, the new technique of Active Optics has been developed for medium or large telescopes, with Image quality optimised automatically by means of constant adjustments by in-built corrective optical elements operating at fairly low temporal frequency ~ 0.05 Hz or less. The first fully active telescope, the ESO 3.5 m New Technology Telescope (NTT), entered into operation at La Silla in 1989. Active optics is very much at the heart of the segmented 10-m Keck primary mirror, in operation since 1992 on Mauna Kea, Hawaii and of e.g. the VLT four 8.2 m thin mirrors, now all operating in Paranal.
To appreciate the daunting task faced by designers of adaptive optics systems, one should understand that an initially plane wavefront travelling 20 km through the turbulent atmosphere accumulates, across the diameter of a large telescope, phase errors of a few micrometers. These have to be sensed with a minimum number of photons and corrected to about 1/50 of a micrometer every millisecond or so. Another complication is that, for short integration times, the field of view over which the atmospheric wavefront distortions and hence the images are correlated, the isoplanatic angle, is very small (only a few arc second for visible wavelengths).
Because of the high bandwidth and the small field to which correction can generally be applied, adaptive optics uses a small deformable mirror with a diameter of 8 to 20 cm located behind the focus of the telescope at or near an image of the pupil. In some current projects, the possibility of using a large deformable secondary mirror is being developed. The choice of the number of (usually piezoelectric) actuators is a tradeoff between degree of correction, use of faint reference sources (see below) and available budget. For instance, a near-perfect correction for an observation done in visible light (0.6 /265m) with an 8-m telescope would require ~ 6400 actuators, whereas a similar performance at 2 /265m needs only 250 actuators.
A large number of actuators requires a similarly large number of subapertures in the wavefront sensor, which means that for correction in the visible, the reference star should be ~ 25 times brighter than to correct in the infrared. Most current astronomical systems are designed to provide diffraction-limited images in the near-infrared (1 to 2 /265m) with the capability for partial correction in the visible. However, some military systems for satellite observations in the USA do provide full correction in the visible on at least 1-m class telescopes.
Two main methods are used to measure the degraded wavefront, the Shack-Hartmann device, which measured the slope of the wavefront from the positions of the images of the reference star given by each subpupil, and curvature sensing, where the intensities measured in strongly defocused images provided directly give the local curvatures of the wavefront. Correction in the Shack-Hartmann device is made with individual piezoelectric actuators. Correction in a curvature sensing system is accomplished with a bimorph adaptive mirror, made of two bonded piezoelectric plates. With both methods, wavefront sensing is done on a reference star, or even on the observed object itself if it is bright enough and has sufficiently sharp light gradients. The measurement can be performed in the visible for observation in the infrared, or in the infrared itself (1 to 2 /265m), if e.g. the reference star is too faint in the visible.
The control system is generally a specialized computer that calculates from the wavefront-sensor measurements the commands sent to the actuators of the deformable mirror. The calculation must be done fast (within 0.5 to 1 ms), otherwise the state of the atmosphere may have changed rendering the wavefront correction inaccurate. The required computing power needed can exceed several hundred million operations for each set of commands sent to a 250-actuator deformable mirror. As in active optics systems, zonal or modal control methods are used. In zonal control, each zone or segment of the mirror is controlled independently by wavefront signals that are measured for the subaperture corresponding to that zone. In modal control, the wavefront is expressed as the linear combination of modes that best fit the atmospheric perturbations.
AO Operation is strongly affected by the size of the isoplanatic angle, usually ~ 20" at 2 /265m, but only ~ 5" at 0.6 /265m. It is generally NOT possible to find a sufficiently bright reference star close enough to an arbitrary astronomical object. Conditions are much better in the infrared than in the visible since atmospheric turbulence (and especially its high spatial frequencies) has, for a given AO correction, a weaker effect on longer wavelengths. The spatial and temporal sampling of the disturbed wavefront can therefore be reduced, which in turn permits the use of fainter reference stars. Coupled with the larger isoplanetic angle in the IR, this gives a much better outlook for AO correction than in the visible.
Nevertheless, even for observations at 2.2 Ám, the sky coverage achievable by this technique (equal to the probability of finding a suitable reference star in the isoplanatic patch around the chosen target) is only of the order of 0.5 to 1%. It is therefore quite normal that most scientific applications of AO so far have been on objects which naturally provide their reference object like solar system small bodies, stellar environments, stellar clusters and a few very bright Seyfert nuclei.
At this time, a number of team or general purposes astronomical AO systems are routinely working on 4-m class or larger telescopes: (COME-ON)ADONIS, the first general purpose AO system, on the ESO-La Silla 3.6 m telescope; (UH-AO)Hokupa'a, the IfA-UH curvature system pioneer, observing at Mauna Kea and on the 8-m Gemini-North telescope; PUEO installed on the 3.6-m CFHT telescope (Mauna Kea); ADOPT on the 100" Hooker telescope (Mount Wilson), ALFA on the Calar Alto 3.5-m telescope, the first to use "routinely" a laser guide star (LGS) projector; the LLNL AO system at the 3.5-m Shane telescope (Lick Observatory), currently with a Natural Guide star only, but soon featuring an LGS; the first AO system on a very large telescope, viz. the Keck II AO facility at Mauna Kea. Many more are under construction or installation, including NAOS and SINFONI for the ESO VLT.
The most promising way to overcome the isoplanatic angle limitation is the use of artificial reference stars, also referred to as laser guide stars (LGS) (Fig. 4). These are patches of light created by the back-scattering of pulsed laser light by sodium atoms in the high mesosphere or by molecules and particles located in the low stratosphere. The laser beam is focused at an altitude of about 90 km in the first case (Sodium resonance) and 10 to 20 km in the second case (Rayleigh diffusion). Such an artificial reference star can be created as close to the astronomical target as desired, and a wavefront sensor measuring the scattered laser light is used to correct the wavefront aberrations on the target object.
Several laboratories in the United States, operating under military contracts, have reported the successful operation of adaptive optics devices at visible wavelengths with a laser guide star on a 60-cm telescope [Defense Advanced Research Projects Agency (DARPA), Maui Optical Station (AMOS) situated on top of Mount Haleakala in Maui, Hawaii] and on a 1.5-m telescope (U.S. Air Force Starfire Optical Range). Both got images with ~ 0.15 arc sec resolution and proved the feasibility of laser probes. A joint program of the Strategic Defense Initiative Organization (SDIO) and the U.S. Navy reported an improved resolution by almost a factor of 10 on a 1-m telescope in San Diego, California. Some systems for astronomical. On the "Civilian" side, the first astronomical observation was done ....., and the Chicago Adaptive Optics System (ChAOS) on the 3.5-m ARC telescope at Apache Point, New Mexico (56).
Nevertheless, there are still a number of physical limitations with an LGS. A first problem, focus anisoplanatism, also called the cone effect, became evident very early on. Because the artificial star is created at a relatively low altitude, back-scattered light collected by the telescope forms a conical beam, which does not cross exactly the same turbulence-layer areas as the light coming from the distant astronomical source. This leads to a phase estimation error, which in principle may be solved by the simultaneous use of several laser guide stars around the observed object. The effect is minimized with the sodium resonance technique and roughly equivalent on an 8-m telescope to the phase error experienced with an NGS 10" away from the astronomical target. This in particular leads to still reasonable performance at 2 /265m with a ~ magnitude 9 beacon.
Even more severe is the image motion or tilt determination problem. Because the paths of the light rays are the same on the way up as on the way down, the centroid of the artificial light spot appears to be stationary in the sky, while the apparent position of an astronomical source suffers lateral motions (also known as tip/tilt). The simplest solution is to supplement the AO system using the LGS with a tip/tilt corrector set on a (generally) faint close NGS. Performance is then limited by the poor photon statistics for correcting the tip/tilt error. A more performant (and complex) solution would be to use two different AO systems with two laser beacons, one, for the astronomical object and one for the reference star. Tip/tilt photon statistics would then be much increased by the star sharpening provided by the 2nd AO system.
With the latter technique, fainter natural reference stars can be used to measure the image motion, so the probability of finding such a reference star close to the astronomical object is higher, This concept of dual adaptive optics therefore provides a better sky coverage (up to 80% for an 8-m telescope at 1- 2 Ám). An obvious implication is that the larger the telescope, the greater the sky coverage because the gain in resolution brought about by the increase of the diameter of the optics is fully exploited. On the other hand, it has severe technological implications, as it requires the duplication of all components (deformable mirror, wavefront sensor, and laser guide star).
Adaptive optics with a multicolour laser probe is another concept investigated to solve the tilt determination problem of laser beacon based AO. Only applicable to sodium resonant scattering at 90 km, it excites different states of the sodium atoms and makes use of the slight variation in the refraction index of air with wavelength. Its main drawback is the limited returned flux, owing to the saturation of mesospheric sodium layer. The multicolour laser guide star may provide corrections without any natural reference star, resulting in a 100% sky coverage, but current tests are not totally encouraging.
The most evident use is direct imaging with filters. All AO systems provide this basic mode, often supplemented with a scanning filter (circular variable filter or scanning Fabry-Perot) to get full data cubes with both the 2D spatial and 1D spectral information on the astronomical targets. Getting these data cubes in a single exposure is very attactive, given the time variable nature of turbulence, even after AO correction. This can be done by the so-called Integral Field Spectrographs (IFS). Their use with AO corrections has been pioneered by OASIS at CFHT in the visible and 3D at Calar Alto for the near-IR. Similar instruments are being developed for the 8-ms, in particular GMOS in the visible at Gemini and SINFONI -SPIFFI in the near-IR at the VLT. GMOS also features a unique multi-slit capability coupled with Adaptive Optics.
There are many substantial technological challenges in AO. Among them are the development of fast, very low-noise detectors in order to be able to correct with fainter reference stars; high-power reliable & easy to operate sodium lasers; very fast processors exceeding 109 to 1010 operations per second; deformable mirrors with bandwidths of several kilohertz and with thousands of actuators, and large secondary adaptive mirrors. The latter are especially interesting at thermal wavelengths, where any additional mirror raises the already huge thermal background seen by the instruments.
NGS-based AO in the Infrared is routinely achieving near diffraction-limited images and spectroscopic data cubes on large telescopes up to the present generation of 8-10 m diameter. Significant corrections have been obtained in the visible in exceptionally good seeing conditions, but diffraction-limited performance has up to now .... Single LGS systems are now or soon operating at a number of Observatories, but routine demonstration of their potential for getting very high sky coverage has not yet been achieved. MCAO techniques are still in their infancy.
Many recent astronomical discoveries can be directly attributed to new optical observation capabilities. With the new generation of Very Large Telescopes entering into operation, the role of AO sytems (and for even better resolution, interferometry) is extremely important. With this capability, their huge light-gathering along with the ability to resolve small details, both spatially and spectrally, has the potential to bring major advances in ground-based astronomy in the new decade. Further down the line, the giant optical telescopes under discussion, like OWL, will rely on advanced AO systems for basically ALL their observations; they will have to be incorporated right at the start of the projects.