After thirty years of space exploration it may come as a surprise that new information on the solar system can be obtained from images made with a ground-based telescope. This is because spacecrafts could only make short term observations with a limited field-of-view. Moreover, until Galileo, space probes were not equipped with infrared cameras. Only recently, the Hubble Space Telescope (HST) has produced sharp images of planets from Earth orbit both in the visible and in the infrared. However, the angular resolution has been limited by the size of the telescope aperture (2.4 m), and observation time has been limited by competition with other programs.

The observations presented here were made with the 3.6-m Canada-France-Hawaii Telescope (CFHT) equipped with adaptive optics. Adaptive optics (AO) uses a "guide" source to sense the image degradation produced by the Earth's atmosphere, and a deformable mirror to compensate it in real time. Most large telescopes are already or will soon be equipped with AO. The results presented here have benefited from the high optical quality of the CFHT (low level of scattered light) and from the use of "curvature" adaptive optics a technique recently developed by F. Roddier and his collaborators at the Institute for Astronomy (IfA) in the University of Hawaii. Compared to conventional AO, curvature AO can operate with fainter guide sources and requires less actuators to deform the mirror.

Click on the images to see them enlarged

Fig. 1. A Voyager image showing the names of Saturn's main rings and divisions. Source: NASA's Planetary Photojournal: Image No. PIA00335.

Fig. 2. A two-dimensional display of Saturn's ring brightness as a function of both distance from Saturn (horizontal axis) and time (vertical axis). Time increases quasi-monotonically from bottom to top. This figure shows data taken on Aug. 10 1995, from 12:15 to 12:56 (UT), that is about 8 hours before the crossing. Vertical bands show the fixed ring structures indicated under the figure. Because the dark side of the rings is seen here, the contrast is inverted, that is transparent divisions which are normally dark (such as the Cassini division) appear bright as they let the sun light shine through. Bright slanted lines are produced by the moving objects indicated on top. Note the temporary disappearance of Epimetheus due to an eclipse in the shadow of the A ring.

The first curvature AO system had only 13 actuators. It was built at the IfA and started to operate in 1994. In August 1995, it was used at the CFHT to observe Saturn as the Earth was crossing the ring plane. As the rings nearly disappear, it is a rare opportunity to observe Saturn's faint satellites otherwise invisible from Earth. This event was simultaneously observed by HST and by the 3.6-m telescope of the European Southern Observatory (ESO) also equipped with an AO system, albeit a conventional one.

Satellites discovered by the Voyager space probe such as Prometheus and Pandora were observed for the first time from the ground. In addition HST observers claimed the discovery of three new objects later identified as clumps in the F ring. Discovered by the Pioneer 11 spacecraft, this faint ring encircles Saturn's main rings. During the crossing, it hides most of them and becomes prominent. Evidence for clumps in the F ring comes from Voyager data.

Not only the CFHT observations revealed most of these objects, but -owing to the exceptional quality of the data- nine additional clumps were also detected (Roddier et al., Icarus, 143, p. 299, 2000). Moreover, two nights after the crossing evidence was found for a decaying arc of particles scattered along an orbit close to that of Enceladus. A possible explanation is that a large block of ice previously ejected by Enceladus collided with ice fragments trapped on the satellite orbit. The collision produced an expanding cloud of small ice particles which quickly vaporized, an event never observed before (Roddier et al., Icarus 136, p. 50, 1998).

Fig. 3. Top figure: same as Fig. 2 except that the distance to Saturn has been converted into longitude, assuming an orbital motion at the distance of the F ring. From bottom to top, the four wide horizontal strips show results obtained with adaptive optics (AO). The thinner strips on top of them are from observations made with the Hubble Space Telescope (HST). In this representation, objects moving at the distance of the F ring produce a straight vertical line, whereas objects orbiting at a different distance produce inclined or curved lines. Note the large number of faint objects orbiting near the F ring. These include objects S5, S6, and S7 announced by the HST team. Both S5 and S7 were also detected with AO, whereas S6 falls in a region that was not observed. Ten additional objects labeled S9 and S11 through S19 where detected with AO. S13 was later confirmed by the HST team. Bottom figure: the location of the clumps on their orbit.

Fig. 4. Evidence for an arc of particles in orbital motion around Saturn. The first five images (a to e) form a time sequence of images with 10-min. intervals. The bottom right image (f) is a median of the five other images. East is left. Janus and the tip of the bright rings are on the right. The arc is the bright horizontal streak moving eastward (away from Saturn). A vertical line (visible in the enlarged image) has been drawn to coincide with the sharp right edge of the arc in the first image (a). Note the foreshortening of the arc as it approaches maximum elongation.

Given the success of curvature AO, the CFHT Corporation built a 19-actuator user AO system based on the same technique. Called after the name of the Hawaiian owl "Pueo", this AO system has been offered to the CFHT users since 1996. A notable discovery made with "Pueo" and already mentioned in the press is that of a moon orbiting asteroid (45)Eugenia (Merline et al., Nature 401, p. 565, 1999). Previous searches for asteroidal satellites with both ground-based telescopes and HST had failed. The only other known such satellite is Dactyl found orbiting (243)Ida by the spacecraft Galileo. The NEAR space probe that was recently put into orbit around EROS has not found yet a companion.

Fig. 5. These false-color images of Neptune were obtained on July 6, 1998 in a methane absorption band (1.72 µm). The top three images are individual 600-second exposures taken at the time (UT) indicated above. Note how the fine structure in the cloud bands can be followed from one frame to another as the planet rotates. These left and right images have been numerically rotated about Neptune's rotation axis to match the central image and co-added to it to form the bottom images, thus improving the signal to noise ratio. Intensity in the bottom right image has been increased to show fainter details. These images have been deconvolved by Thierry Fusco ( with the MISTRAL (1) algorithm, a deconvolution algorithm which preserves sharp edges.

(1) MISTRAL : Myopic Iterative STep-preserving Restoration ALgorithm.

In 1997, a 36-actuator curvature AO system was built at the IfA. Called "Hokupa'a" after the Hawaiian name of the North star, this visitor AO system was mounted on the CFHT and used to observe Neptune both in November 1997 and in July 1998. It produced the first sharp infrared images of Neptune. These images show the fine structure of its cloud bands with high contrast, allowing the details of Neptune's atmospheric activity to be monitored from the ground for the first time (Roddier et al., Icarus 136, p. 168, 1998). Deep images taken in July revealed four of the dark satellites discovered by Voyager 2. In addition to Proteus previously detected before, Larissa, Galatea, and Despina were detected for the first time from a ground-based telescope.

Moreover, evidence was found for arcs at the distance of the Adams ring (Sicardy et al., Nature 400, p. 731, 1999). Similar arcs were discovered by Voyager 2, but an ambiguity remained on their motion leaving two possible orbital solutions. From theoretical considerations one solution was considered more likely than the other. Curiously, the ground-based observation (confirmed by HST observations) was found to be consistent with the less likely solution, calling current theory into question. Ground-based observations also showed evidence for arcs in the Leverrier ring (IAU circular 7108), a possible transient phenomenon since no such arcs were seen by Voyager 2.

Fig. 6. Deep (false-color) composite image of Neptune showing faint orbiting objects. Top : as observed. Bottom: showing the orbits of Proteus, Larissa, Galatea (light blue lines), and of the Le Verrier and Adams rings (white lines). Data were taken through a narrow-band filter centered on a methane absorption band (1.72 µm) where Neptune is darker, to minimize light scattered by the telescope optics. In the computer processing, Neptune itself was masked (dark area in the middle) to further minimize contamination. The three satellites appear on the left side as red spots, with a slight elongation due to their orbital motion during the 600-second exposures. The Adams ring arcs are seen at western elongation (right side of Neptune), just outside Galatea's orbit. Note also the excess of brightness along the Le Verrier ring's orbit, only on the right side of Neptune. This could be due to a temporary arc never observed before.

After July 1998, the Hokupa'a system was rebuilt at the IfA to fit the 8-m Gemini telescope, where it produced the images widely reported in the press at the June 1999 telescope dedication. This system is now being modified to accommodate 85 actuators. Given the success of curvature AO systems for planetary observations, NASA has decided to equip its Infrared Telescope Facility (IRTF) on Mauna Kea with a 36-actuator curvature AO system. Similar to the current version of Hokupa'a, this user instrument is currently being built at the IfA.

The development of curvature AO as well as the construction and upgrade of Hokupa'a was made possible by continuous funding from the National Science Foundation (NSF) since 1991. The IfA program for AO observations in planetary science was supported by NASA. CFHT is funded by the National Research Council (NRC) of Canada, the Centre National de la Recherche Scientifique (CNRS) of France, and the University of Hawaii.

More results from the Adaptive Optics group at the Institute for Astronomy can be found on : .

Contact :
Francois Roddier - Institute for Astronomy - University of Hawaii
Phone : (808) 956 8429
E-mail :

More information on the Canada-France-Hawaii Telescope can be found on :
Contact :
Christian Veillet, senior resident astronomer at CFHT
Phone : (808) 885 3161
E-mail :