eso9006 — Organisation Release
Adaptive Optics sharpens telescopes' sight
23 May 1990
With the help of "adaptive optics," a revolutionary optical concept (eso8908), infrared astronomical images have been obtained with the ESO 3.6 m telescope at the La Silla observatory which are as sharp as they would be if the telescope were situated in space. This is the first time in astronomy that a ground-based telescope of this size has been able to directly register during long time periods stellar images with a sharpness that corresponds to the theoretically possible limit.
Producing four times sharper images than possible before, the 3.6-m telescope is now able to register images up to sixteen times fainter than before. With the new technique its observational potential in the infrared spectral region is unsurpassed by any other ground- or space-based telescope.
The observations were made by an astronomer/engineer team with a new device, the VLT Adaptive Optics Prototype System, developed in a collaboration between ESO, the Office National d'Etudes et de Recherches Aérospatiales (ONERA), LASERDOT (formerly Laboratoires de Marcoussis) and the Observatoire de Paris-Meudon in France. The principle of adaptive optics is based on a computer-controlled, small deformable mirror which counteracts the smearing effect of the atmospheric turbulence. More details may be found in the Appendix at the end of this Press Release.
Eliminating the atmospherical turbulence
During a period of good observing conditions, images were obtained in various wavebands in the infrared region of the spectrum, ranging from the J-band (at wavelength 1.2 μm) to the M-band (4.8 μm). The atmospherically induced "seeing", i.e. the diameter of stellar images registered by the telescope, was around 0.8 arcsec; the actuation of the adaptive optical system reduced this by a factor of four in the L-band (3.5 μm), to the theoretically smallest possible value, 0.22 arcsec. The picture accompanying this Press Release illustrates the dramatic improvement in image sharpness . With the adaptive optics, the telescope easily separates the components of a double star which are at a distance of only 0.38 arcseconds.
This achievement implies that the image-smearing effect of the turbulence in the atmosphere above the telescope was almost completely eliminated. It no longer has any influence on the image sharpness and for observations at this wavelength, the 3.6 m telescope therefore functioned as if it were situated in space.
The present prototype adaptive optics system is optimized for this wavelength region, but a substantial improvement was also seen at shorter wavelengths. For instance, the image diameter in the K-band (2.2 μm) was measured as 0.18 arcseconds. This is the smallest image size ever obtained continuously and in real-time at a large ground-based telescope and it is only slightly above the theoretical limit in this waveband, 0.13 arcseconds.
Following the successful demonstration of the adaptive optics principle at the 3.6-m telescope, the team of astronomers and engineers at ESO and in France now aims at the rapid implementation of further technological refinements in a second-generation adaptive optics instrument.
By increasing the number of computer-controlled supports ("actuators") for the deformable mirror from 19 to 52 and improving the speed of computation of the mirror corrections by a factor of 2.5 or more, it is expected that it will become possible to achieve the theoretical sharpness limit at shorter wavelengths, e.g. in the H-band (1.65 μm). This would also further improve the performance at even shorter wavelengths; the present configuration already reduces the image diameters in the J-band (1.2 μm) by a factor of about three, to 0.31 arcseconds, while the theoretical limit at this waveband is 0.07 arcseconds. It is expected that the new developments will require about 18 months.
Not only does adaptive optics increase the image sharpness, it also concentrates the light better (see the picture). This is of great importance in order to increase the efficiency of the modern detectors at the telescope.
There is little doubt that adaptive optics will play an increasingly important role in ground-based astronomy. The elimination of the adverse effects of atmospheric turbulence enables a ground-based telescope to approach - in the restricted wavelength range where the Earth's atmosphere is transparent, i.e. mostly at near-infrared wavelengths - the limits of a space-based one, but at a much smaller cost. This technique is therefore entirely complementary to the Hubble Space Telescope concept.
Adaptive optics will provide a decisive advantage for the interferometric mode of ESO's 16-m Very Large Telescope (VLT), a unique feature of this project that combines several telescopes.
But are there limitations? Yes, the adaptive optics principle only works on a small sky field around a comparatively bright reference star. The light from this star gives the information that is needed to control the deformable mirror.
Appendix: what is adaptive optics?
Ever since the invention of the telescope in the early 17th century, astronomers have had to accept that the sharpness of astronomical images obtained with ground-based instruments is severely limited by a factor which is beyond their control, that is the turbulence in the Earth's atmosphere.
This turbulence is perceived by the eye as the twinkling of stars. High above the observer, mostly at altitudes between 5 and 10 kilometres, there are many small, moving cells of air, each of which produces a "sub-image" of the same star; the result is a swarm of moving sub-images. (Compare with the air above a toaster or a hot radiator.)
To a naked-eye observer, the number of sub-images which fall within the periphery of his eye pupil changes all the time. The perceived intensity of the star varies; the star twinkles.
In a telescope, the size (that is, the sharpness) of a stellar image, is equal to the area within which this swarm of sub-images moves. The greater the air turbulence, the larger is this area and the less sharp are the resulting images. Because of this effect, an increase of the size of a telescope does not improve its ability to resolve details of astronomical objects, once the aperture of the telescope exceeds 10 or 20 cm; the best achievable image sharpness, even by high-quality, large astronomical telescopes, is effectively determined by the state of the atmosphere, and is referred to as "astronomical seeing" during the exposure. For this reason, large telescopes are placed at sites where the atmospheric turbulence is as small as possible, for instance La Silla.
The technique of "adaptive optics" overcomes this natural limit. Expressed in simple terms, it enables the telescope to "catch" all of the subimages by means of a small, deformable mirror which "focuses" these images into one, sharp image. It can also be described in terms of correcting the atmospherically introduced distortions of the light wavefront from the star.
It is based on a feed-back loop, and the optical system contains a deformable mirror which can change its surface profile in a way that exactly compensates for the distortions of the light wavefront after it has passed through the atmosphere. The information about how to deform the mirror comes from a wavefront sensor which allows to measure the shape of the distorted light wavefront. It requires a very fast and powerful computer to calculate how the actuators located behind the deformable mirror have to push and pull the mirror surface.
The present prototype system has a mirror with 19 actuators. The mirror is deformed, hence the wavefront is corrected, 100 times per second.
Note that the technique of adaptive optics, as described here, is complementary to active optics, a system that allows to keep large astronomical mirrors in optimal shape when gravity, wind and heat distort them, and which has been so successfully implemented at the ESO New Technology Telescope (see eso8903, eso8904 and eso9003).
Members: Gerard Rousset (ONERA); Fritz Merkle and Georg Gehring (ESO); Francois Rigaut, Pierre Kern and Pierre Gigan (Observatoire de Paris); Corinne Boyer (LASERDOT).
Tel: +49 89 3200 6276