Report of the Working Group on Satellites for the Period July 1987 - June 1990 (J.E. Arlot) 1 Observations for Astrometric Purposes 1.1 Photographic observations USNO (Washington D.C., USA): Pascu and Schmidt ( Astron.J., 99, 1974, 1990) continued observations of the Galilean satellites, the satellites of Mars and S I-VIII with the 26-in refractor in Washington D.C.. The satellites of Uranus and Neptune (N I) were observed (Walker and Harrington 45.101.072). Observatorio Nacional - LNA (Itajub\'a, Brazil): positions from 264 plates of the satellites of Uranus (1982 - 1985) with the 1.6m-reflector were published (Veiga et al.44.101.017). 37 plates of the satellites of Saturn, 74 plates of N I, 2 plates of N II were obtained and will be reduced. Kiso Observatory (Japan): positions from 46 plates of J VI, VII, VIII and IX (1986 - 1990) with the 105cm Schmidt were partially published (Nakamura et al. 45.099.029). ESO (La Silla, Chile) and CFH (Hawaii, USA): positions of N II from observations in 1982 and 1984 with the 1.5m-Danish reflector at ESO and in 1982 with the 3.6m-CFH reflector were published (Veillet and Bois45.101.005); positions of the lagrangian satellites of Thetys and Dione were similarly obtained in 1981 and 1985 (Oberti et al.50.100.028). Bordeaux Observatory (France): photographic observations of the satellites of Saturn made in 1984 (Dourneau et al.50.100.075). Astrometric observations of Titan and Iapetus are made in the FK5 system with the automatic meridian circle at each opposition since 1985 (60 for Titan and 25 for Iapetus until 1990); they have not yet been published. La Palma Observatory (Canary Islands, Spain): photographic positions of N I were obtained with the 1m-reflector (Taylor et al. A&A, 232, 565). USSR Observatories: Mars: In view of the importance of determining accurately the orbit of Phobos for the Phobos space project, an observational campaign was organized in 1988. More than ten observatories within the USSR participated and two expeditions were sent abroad; the observations were processed and transmitted to the Space Navigation Center and other institutions for improvement of the orbits. 856 positions of Phobos and 937 positions of Deimos (r.m.s. = 0.12 arcsec) were obtained at Majdanak during 19 nights (July 23 - Oct.4, 1988) by staff members of GAISh (Moscow) (RZh.8.51.140, 1989; RZh.12.51.255, 1989) and in 1986 by the Goloseevo expedition (RZh.4.51.75, 1989). Observations of Phobos were made in 1986 at Abastumani (Kiseleva and Chanturija 46.097,150) and long series of observations were also obtained at Pulkovo, Ordubad and other observatories. Jupiter: Observations of the Galilean satellites continued at Pulkovo with the 26-inch refractor and the normal astrograph (Kiseleva 45.099.007); 131 plates of the satellites were taken. At Abastumani 64 plates were obtained and 36 positions were determined at the Nikolaev. Observations of these objects were also made in Kitab, Tashkent and Goloseevo; 143 positions determined at Nikolaev in 1983 -1985 were published (Voronenko and Gorel' 46.099.056), 95 further positions (1986-87) are in press. Observations at Abastumani (1983-84) and at Ordubad were published (Kiseleva et al.45.041.023, Bobylev and Dement'eva 46.099.116). Saturn: Observations of the satellites of Saturn were made at Abastumani (89 plates), at Goloseevo, in Kazan (33 plates) and Tashkent. Results of observations in 1976 with the 26-inch refractor at Pulkovo, in 1983-85 with the zone astrograph in Nikolaev, and of S II - S VIII in 1983-84 at Abastumani have appeared (Tolbin 46.100.067; Voronenko and Gorel' 46.099.056; Kiseleva et al.45.041.023). Uranus and Neptune: At Abastumani 91 plates with U III and U IV, and 83 plates with Neptune and N I were taken. 1.2 CCD observations: USNO (Washington D.C., USA): J XIV, S XII, S XIII, S XIV, U V and N II were observed with the 61-inch reflector at Flagstaff; positions of U V (1981 - 85) were published (Pascu et al.43.101.025) and positions of N II were sent to JPL for the Voyager mission at Neptune. La Palma Observatory (Canary Islands, Spain): positions of N I were obtained with the 1m-telescope (Taylor et al. A&A, 232, 565). Pic du Midi (France): during the 1988 Mars opposition, observations of Phobos and Deimos were made using the 1m telescope and 813 positions were obtained (Colas and Arlot: in press). 1.3 Photometric observations: Mutual events of the Galilean satellites were observed in 1985 (Froeschle et al.45.099.001; Melillo 44.099.051; Arlot et al.49.099.010; Arlot et al.50.099.085, Arlot et al. A&A Supp.Ser. 82, 513, 1990) and the timing of jovian eclipses was studied by Loader (45.099.013) while reports of observations were made by Westphall (44.099.038). Mutual events of Pluto-Charon were observed by Vasundhara and Bhattacharyya (45.101.077) and by Blanco et al.(50.101.129). The occultation of 28 Sgr by Titan gave accurate astrometric results from observations in Paris, Meudon, Pic du Midi and Catania (Sicardy et al. Nature 343, 350, 1990), at the Vatican and in Israel (Hubbard et al. ibid, 353) as well as in the USSR ( Odessa, Kharkov, Uzhgorod, Kiev, Crimea) and in Western Europe (cf. section V.3). 1.4 Other observations: 37 observations were made on-board Phobos-2 on Febr.21, Febr.28 and March 25, 1989, at a range from 1100 to 180 km. Processing of those made in February contributed to successful spacecraft trim maneuvers. As a result, Phobos-2 was inserted into a ``quasi-satellite'' orbit (relative to Phobos) and continued to stay in the closest vicinity of Phobos. Astrometric observations of Phobos and Deimos from Mariner 9 (1971-72) were published by Duxbury and Callahan (49.097.018). 56 speckle interferometric observations of Charon were made in 1984-85 by Beletic et al.(49.101.018). During the triennium, two compilations of observations were published: a catalogue of 5767 ground-based astrometric observations of the satellites of Mars made during the period 1877-1982 (Morley49.002.012) and a catalogue of 51000 observations of S I - VIII made during the period 1874-1989 (Strugnell and Taylor: A&A Suppl.Ser., 83, 289). 2 Comparison of Observations with Theories -Determination of Elements 2.1 Satellites of Mars: The Phobos space project actively encouraged studies of the motions of the Martian satellites. Subsequent improvement of the orbits was achieved by means of the new ground-based observations in 1988 and also from previously unpublished observations by Pascu in 1969 and 1971. The use of the many, high-precision observations resulted in greatly improved agreement between orbits determined from ground-based observations and from on-board observations, respectively (Shor ACM III 175). The 1-sigma uncertainty of the predicted positions of Phobos was reduced to 5 km (Kudryavtsevet al. Preprint GAISH N2, Moscow, 19, 1989). As mentioned above, the improved orbital elements for Phobos were used for trim maneuvers of Phobos-2 and for on-board observations of Phobos, which immediately proved the high accuracy of the Phobos ephemeris; the (O-C) values did not exceed 3 km (Kudryavtsev et al., ibid). On-board observations of Phobos combined with telemetry data made it possible to improve the orbit, to facilitate the spacecraft navigation and to determine Phobos' mass from the gravitational perturbations of the spacecraft. The uncertainty of the position of Phobos is now considered to be equal to 2-3 km (Kudryavtsev et al., ibid). The observational data have been analyzed by Shor (46.097.049), Ivanov et al.(46.097.049), Jones et al.(49.097.016), Sinclair (50.097.023), Jacobson et al.(50.097.054), Morley (49.002.012; AAS/GSFC Int.Symp. on Orbital Mechanics and Mission Design (Greenbelt), Paper 89-181, 1989) and Chapront-Touze ( A&A, 235, 447, 1990), in order to determine the orbital elements of the satellites, various physical parameters of Mars, and the secular acceleration of Phobos. In addition, extensive analyses (some unpublished) have been carried out by the space mission engineers of Intercosmos/Glavcosmos, ESA/ESOC and NASA/JPL in support of the Phobos mission. There is agreement that a secular acceleration of Phobos exists: the well-determined value from the observations is $0.00124 pm 0.00002$ deg yr$^{-2$, which in fact is not too far from Sharpless' original (1945) value of 0.00188. It is now clear that the uncertainty of the value of the acceleration generated by the disparity of various determinations was caused by a misinterpretation of the time-scale of some early observations from the Lick Observatory in a widely-used listing, and further confounded by an error in the time-scale of an early release of the Mariner 9 data. 2.2 Satellites of Saturn: The orbital elements of S I - VIII were improved by means of 14000 photographic observations (1967-83), emphasizing the need to use the best available theories for Hyperion and Iapetus (Taylor and Chen46.100.001). An analysis of the orbits of Titan, Hyperion and Iapetus by numerical integration with a fit to micrometer observations during 1873-1923 was also made (Harper et al.50.100.029). Photographic observations made in 1971 were used for improvements of the orbital elements of Iapetus and Hyperion (Hatanaka45.100.013) and observations of the Saturnian system made in 1975 were compared with the theories of motion (Tolbin44.100.030). From an analysis of Voyager 1 and 2 data and ground-based observations obtained during the 1966 and 1980 ring plane crossings, a determination of the orbits and masses of 1980 S X (Janus) and 1980 S XI (Epimetheus) was made (Yoder et al.50.100.085). 2.3 Satellites of Uranus: Numerous works appeared during the past triennium because of the Voyager encounter with Uranus. After the encounter, JPL published observations as well as various dynamical studies (44.003.001). Results on the masses of U I - V were also published by Anderson et al.(45.101.018). An analytical ephemeris of U I - V was deduced from the Laskar's GUST86 theory (Laskar and Jacobson44.101.017). A comparison of ground-based observations with theory was made for 1982-85 observations (Veiga et al.44.101.017). Batrakov and Nikolskaya (50.101.053) published improved orbital parameters for U I - V from photographic observations in 1968-86; the observations are reproduced with r.m.s. values $sim 0.2$ to 0.3 arcsec. 2.4 Satellites of Neptune: Orbital elements was determined from observations during 1982-84 (Veillet and Bois45.101.005). Ephemerides used by the Voyager project were based on orbits determined by Jacobson ( A&A, 231, 241). 2.5 Satellite of Pluto: Orbital elements of Charon were determined from speckle interferometric observations in 1984 and 1985 (Beletic et al.49.101.045) 3 Theoretical Studies A review paper on the long-term evolution of the orbits of natural satellites was published by Duriez (46.091.087). Other studies of the evolution of natural planetary satellite systems were published by Peale (50.107.061) and Horedt (49.107.002). 3.1 Satellites of Mars: Morley ( A&A, 228, 260) improved Sinclair's analytical orbital model of the satellites to an accuracy of about 100 m, in order to be compatible with the accuracy of the best spacecraft observations of the satellites ($sim 1$ km). Chapront-Touze (46.097.056; A&A 235, 447) has developed a new semi-analytical theory ESAPHO of the motion of Phobos, which aims at an accuracy of about 1 m, in order to be compatible with the high accuracy of the tracking data that had been expected from the Phobos-lander. 3.2 Satellites of Saturn: Sinclair's theory of Iapetus has been compared with Sinclair and Taylor's numerical integration and the theory of the motion of Iapetus has been improved (Harper et al.45.100.009). A study of periodic orbits in the Enceladus-Dione system has been made (Bevilacqua et al. ODNAO, 13). A theory of the motion of the lagrangian satellites of Thetys and Dione was established by Oberti ( A&A, 228, 275). Message (50.100.092) gives some computer-based techniques which have been developed in the construction of the long-period perturbations of Hyperion's orbit. A general theory of the motion of the first eight satellites of Saturn by Duriez and Vienne is in press ( A&A). 3.3 Satellites of Uranus: Theoretical work continued during the past triennium, mainly because of the Voyager encounter with Uranus. A semi-analytical solution for the eccentricities and longitudes of the pericenters of U I - V was made (Lazzaro et al.44.101.025) and Lazzaro ( ODNAO, 39) studied the effects of solar perturbations on the motion of U I-V. A study of the origin of the chaotic behaviour in the Miranda-Umbriel 3:1 resonance was made by Henrard and Sato (1989, Celest. Mech., vol.47, p.391). Studies were also made of the tidal evolution of the satellites in three papers by Tittemore and Wisdom: I: the passage of Ariel and Umbriel through the 5:3 mean motion commensurability (45.101.074); II: an explanation of the high orbital inclination of Miranda (49.101.028); III: the evolution through the Miranda-Umbriel 3:1, Miranda-Ariel 5:3 and Ariel-Umbriel 2:1 mean motion commensurabilities ( Icarus, 85 394, 1990). The secular perturbations of the satellites were investigated by Malhotra et al.(50.101.052) and the role of secondary resonances in the orbital history of Miranda by Malhotra and Dermott (1990, Icarus, 85, 444). 3.4 Satellites of Neptune: Several studies were made concerning the perturbation of the motion of Nereid (Vieira-Martins ODNAO, 59); Alfimova and Gerasimov44.101.050). An analytic modelling of the motion of Nereid was made by Oberti ( A&A, in press); the tidal evolution in the Neptune-Triton system was discussed by Chyba et al.(50.101.039). 4 Research on Ephemerides - Predictions of Phenomena For the Voyager Neptune encounter, ephemerides of N I-II were prepared by Jacobson (46.101.025). Post-encounter ephemerides of N I-II and 1989 N1 were published by Jacobson et al.(AIAA Paper 90-2881, 1990). The occurrence of mutual phenomena of Galilean satellites in 1991 has led to the publication of several predictions of these events (Aksnes and Franklin Icarus, 84, 542, 1990; Arlot A&A, in press) and a special study was made for the occultations of Io, cf.Sect.V.3. Circumstances for Pluto-Charon mutual events in 1988 were published by Tholen et al.44.101.038). 5 Dynamics of the Ring Systems Jupiter's rings have been described by Showalter et al.(43.099.058), eccentric features were studied in the C ring of Saturn by Porco and Nicholson (44.100.024) and regular structures are displayed in the Cassini Division by Flynn and Cuzzi (50.100.080). Eplee and Smith (43.100.013) explored the spokes in the B ring of Saturn. Cuzzi and Burns (45.100.014) proposed the existence of a belt of moonlets to explain a depletion of charged particles around Saturn's F ring. Rosen and Lissauer (46.100.012) deduce the properties of the Saturnian rings from a bending wave excited by Titan. Lissauer et al.(46.100.064) give a bombardment history of the Saturn system. Uranus' rings have been analyzed by comparing the Voyager 2 data with ground-based observations of occultations by French et al.(45.101.008) for the visible wavelenghts, by Holberg et al.(44.101.026) for the UV, and by Gresh et al.(49.101.031) for the radio data. A model of formation of the Uranus rings was proposed by Esposito and Colwell (49.101.029). Normal modes of oscillation in narrow rings were studied by Papaloizou and Lin (46.062.054) and for the Uranian rings by Longaretti (50.101.150). Borderies and Longaretti (44.091.046) describe the dynamical behaviour of planetary rings in term of streamlines. Meyer-Vernet and Sicardy (43.091.009) analyze the torque exerted by a satellite on a disk at the Lindblad resonances, while Borderies et al.(50.091.015) study the confinement of sharp edges by shepherd satellites. As shown by Greenberg (46.091.010), the physical properties of the particles are a key problem of ring dynamics and Wiesel (44.091.018) shows that some fragmented, narrow ringlets may be explained by very inelastic collisions. Lin et al.(44.042.010) propose a model for the confinement of planetary arcs. A numerical approach is used by Brophy et al.( Icarus, 83, 133, 1990) to present the evolution of a narrow two-component ring with different size particles and by Petit and Henon (44.091.041, 45.091.046) to study the mass segregation and the confinement mechanism. Simulations of colliding particle rings are made by Salo (43.042.057) and by Wisdom and Tremaine (45.091.023). A continuing effort is being made to observe Neptune's ring-like arcs from the ground by means of stellar occultations. Observational constraints and theoretical models of Neptune's arcs have been reviewed by Brahic and Hubbard (49.101.058) and by Lissauer and Nicholson ( Adv. Space Res., 10-1, 231, 1990).