Successful experience of applying the Chebyshev polynomials as a power “mathematical tool” for numerical integration and approximation techniques in celestial mechanics is presented. Detailed analysis of approximation function behavior inside an integration step allows to elaborate a special technique for high accuracy and rapid integration of piece-wise continuous functions, modeling the Earth's shadow effect for artificial satellite orbits. Original software is elaborated for creating the ephemeris file simultaneously with the process of numerical integration. This technique is applied for the construction of ephemerides of natural and artificial celestial bodies as well as for the compact polynomial representation of different geodynamic parameters.

CERES is a powerful software package for calculating the ephemerides of the major planets, minor planets and comets and for executing related tasks.

To make a photograph of a satellite, we have to know how it is located relative to the planet. After processing the plates, the identification of satellites is to be made. Precise coordinates of natural satellites need to be known for this purpose.

La puissance de calcul des ordinateurs PC actuels permet leur utilisation pour des travaux qui étaient, jusqu'à ces dernières années, effectués exclusivement sur de gros ordinateurs. Le mode interactif utilisé pour la gestion et la saisie des données, facilement mis en œuvre sur les PC, permet un gain de temps important par rapport au travail classique en traitement par lot (batch). Nous présentons ici un logiciel PC, permettant le calcul et l'amélioration des éléments orbitaux des comètes. Une bonne connaissance de ces éléments est indispensable pour le calcul des éphémérides, notamment pour la redécouverte des comètes périodiques.

Pour calculer les éphémérides nécessaires à l'observation du Soleil, publiées dans la Connaissance des Temps (CdT), les astronomes du Bureau des longitudes ont utilisé depuis 200 ans différentes valeurs de son diamètre: à l'apogée ou à la distance moyenne de la Terre. Ces données, provenant des meilleures mesures les plus récentes, ont évolué avec la précision des techniques instrumentales.

Groundbased radar has considerable astrometric potential for asteroids, comets and natural satellites that enter the detectability windows of available telescopes. For very closely approaching asteroids and comets, measurements of the distribution of echo power in time delay (range) and Doppler frequency (radial velocity) can achieve a fractional precision between 10−5 and 10−9, and consequently are invaluable for refining orbits and prediction ephemerides. Even for mainbelt asteroids and the more readily detectable planetary satellites whose orbits are very well known, radar can collapse range uncertainties from a few target radii to a few percent of a target radius, with direct implications for the navigation of spacecraft on flyby or rendezvous trajectories.

CCD detectors are rapidly replacing the photographic plate and photomultiplier in satellite observations used for orbital improvement. This includes both phenomena timings as well as tangent plane astrometry. In most cases this change has been for the better, but in some areas there has been no gain – even a loss. We will review this change in terms of the recent history of satellite observations. The impact of the CCD will be discussed in terms of its applications, and the increase in precision it affords. Finally, a few things will be said about future directions, especially about spin-off applications.

Since beginning an automated full-time survey for Near-Earth objects in September 1990, the Spacewatch project has discovered 95 new Near-Earth asteroids (NEAs), 3 new comets, and 3 new Centaur asteroids. Spacewatch typically identifies about 2000 main-belt asteroids each lunation while covering about 150 square degrees to a limiting magnitude of Vlim ∼ 20.9. We report automatically measured astrometric asteroid detections to the Minor Planet Center where known and multiply detected objects are identified. NEAs and other interesting objects are identified by their angular rates of motion near opposition at the time of discovery and are scheduled for astrometric follow-up on subsequent nights. Objects with exceptionally high rates of motion, called very fast moving objects, have been detected in near real-time by the observer and followed for several hours to several days. These objects are the smallest yet detected outside the Earth's atmosphere. Careful analysis of their discovery rates and orbits have indicated an enhancement of their magnitude-frequency distribution over that anticipated before the Spacewatch survey began – of about a factor of 40 for objects near absolute magnitude H ∼ 29 (Rabinowitz, 1993; 1994). A subset of these small objects which have almost circular orbits and perihelia near the orbit of Earth have been recognized as having significantly different orbits from those of the previously known NEAs (Rabinowitz et al., 1993). Their origin is still under debate, with possible sources including Earth or Lunar impact ejecta, Earth-Sun Trojans, or more complicated secular resonance interactions of NEA orbits with the giant planets combined with stochastic perturbational encounters with the inner planets (Bottke, 1994; N.W. Harris, 1995, personal communication). The large volume of asteroid detections allows magnitude–frequency studies of the detected main-belt asteroids and Jupiter Trojans. New discoveries of Comets and Centaur asteroids (whose orbits cross those of the outer planets) may allow studies of their magnitude–frequency distributions as well.

The Carlsberg astrometric telescope has made about 17000 observations of outer Solar System objects since it began operation in 1984. The observed positions of the major planets are compared with JPL DE200 and DE403. The agreement with DE403 is good in general, but unresolved discrepancies of the order 0″.1 are found in Jupiter and Saturn. The run-off between the observations and DE200 which was fitted to observations before 1980 emphasize the need to continue optical observations of the outer planets.

The PLANETS database has been conceived and firstly compiled at the Golosiiv Observatory in 1984, Kiev (Kharin et al. 1987). Since 1988, it is updated, maintained and analysed in cooperation with the Institute of Astronomy, Moscow. By this time the database comprises most of the published optical observations of the Sun and 7 major planets made from 1960 onwards with 21 meridian instruments, 15 astrographs and 11 astrolabes at 29 observatories.

Lunar laser ranging (LLR) (Dickey et al., 1994) consists of measuring changes in the round-trip travel time for a laser pulse traveling between a transmitter on the Earth and a reflector on the Moon. The lunar surface reflectors are still operating normally after almost three decades of use. The ranging data exhibit a rich spectrum of change due to many effects.

A future Russian astrometric satellite (Struve) is aimed at an extension of the fundamental coordinate system down to 18-th magnitude stars. The positional accuracy of observations is planned to be about 0.6 mas. Many of the Solar System objects will be observed by the Struve satellite. Problems of registration and reduction of these observations are discussed.

We present a method to obtain the reference system for isolated observations of faint satellites made with CCD1. The method consists in the construction of a secondary catalogue of faint stars using ‘The Digitized Sky Survey’ and ‘The Guide Star Catalogue’ corrected by modern astrometric catalogues.

This paper reports on the international PHESAT95 campaign of observations of the Saturnian events coordinated by Bureau des longitudes. Thanks to CCD or photometric receptors, accurate astrometric data can be get from the observation of the eclipses by Saturn and mutual events of the Saturnian satellites. These events occur from 1994 to 1996 and we give our first results.

Photographic observations of the Galilean satellites of Jupiter have been made in Pulkovo with 26-inch refractor during two periods 1976–1981 and 1986–1993. The internal mean square errors of jovicentric coordinates and mutual distances is equal to 0.10 arcsec. The external errors are equal to 0.08 and 0.17 arcsec in AR and Decl. The systematic errors were analysed. The errors of the theory G-5 of J.E. Arlot do not exceed 0.1 arcsec.

Astrometric accuracy has two components; the accuracy with which an image can be centred on a CCD or photographic plate and the accuracy with which two image centres can be mapped on to standard co-ordinates. Photographic plates cover a sufficient area of sky that several standard stars can be measured with the satellite images and the mapping to standard coordinates can be done in one operation. CCDs have much smaller areas and the number of stars with sufficiently accurate positions may not be sufficient. The surface densities of the best available catalogues are shown in Table I.

The Charge Coupled Device (CCD) has become the detector of choice in astrometry. The mechanical or electronic (drift scanning) tracking of objects by the telescope leads among observing strategies developed for CCD observations (Stone and Dahn, 1995).

The ESA satellite Hipparcos provides valuable photometric and astrometric results on minor planets and natural satellites. Observations of 48 asteroids, JII-Europa, S VI-Titan and S VIII-Iapetus were made from November 1989 to March 1993. A twenty seconds averaged normal place is constructed, providing thus positions accurate to a few hundredths of arcseconds and relative to a very precise and homogeneous reference frame.

For the determination of the dynamical equinox in fundamental star catalogues observed by astrometric satellites it is proposed to observe occultations of stars by selected asteroids. Being a long term programme now is the time for planning. Quadrature observations and photometric back-up are essential.

In order to evaluate a possible rotation between the Hipparcos and the dynamical reference frames, Hipparcos minor planets preliminary data are analysed. The resolution of the problem is very sensitive to correlations induced by the short length of the interval of observation. Several statistical methods are performed to appreciate the factors of bad conditioning. A procedure for variable selection and model building is given.