The modern crowded CCD-observations of the Solar system small bodies are very effective and accurate. They easily provide close positions on a short topocentric arc for any celestial body – from several hours during one night to several successive nights at a single observatory. It is obvious, now, that the informational value of such short arc observations is very high. Statistical treatment of these accurate near positions gives an opportunity to calculate the first and second derivatives of spherical asteroid coordinates or an angular topocentric velocity and its positional angle, a topocentric angular acceleration and a curvature of visible trajectory of observed celestial body.

Mercury transits are important for the investigation of long-term variations in the Earth rotation. They have been observed for more than 300 years. The basic array of this set is the visual observation of contacts (as a rule the second and third ones). The detection of the instant corresponding to the geometrical contact of the solar and Mercury limbs is a difficult task, since during 10s to 60s, the set of phases is observed transforming continuously from one to another. One of the encumbering factors, is the so-called “black drop” phenomenon, i.e., the dark cross-bar formed between the limbs (Struve, 1882; Kuhl, 1929; Wittman, 1974; Morrison et al., 1975). The influence of this factor could be negleted if the observations were distributed uniformly with regard to contacts, observers and observational conditions. In fact, it is far from that and the necessity of reduction corrections arises.

After a short introduction on the definition and the construction of a reference frame, the presentation concentrates on the actual construction of the new IAU conventional reference frame. The 1991 IAU resolutions on reference systems is reviewed and the present state of the implementation of these resolutions is discussed. The IERS extragalactic reference frame is described and we show that it fulfills the IAU requirements. The Hipparcos catalogue will be linked to this frame before the end of this year. The state of advancement of this job is presented.

The ongoing movement of standardization in Fundamental Astronomy was reviewed. Its history was briefly presented with an emphasis on the problems which triggered its creation. The achievements of the first term of the IAU WG on Astronomical Standards were given. The major goals of a second term were presented with the author's view to resolve them.

Minor planets optical observations have long been used for the purpose of establishing a Celestial reference frame. Being in existence since the early 1960s modern high-accuracy radar measurements of the so-called near-Earth asteroids (NEAs) have been widely extended to the orbit determination process and predicting of the next apparition of the asteroid. Even few radar measurements, when added to optical ones, significantly improve asteroid's ephemeris and reduce standard deviations of the orbital elements (Yeomans et al., 1987). The idea to connect optical and radar data in the problem of the catalogue zero-point determination has been stated by several scientists (Boiko, 1975). And even the first attempt of the authors (Krivova et al., 1994) with actual optical and radar observations of two NEAs: (4179) Toutatis and (1862) Apollo appears to have considerable promise. It was demonstrated the possibility of obtaining standard deviations of catalogue orientation parameters 1.5–2 times better with radar data included.

Several methods of data analysis applied recently to precise astrometric observations give evidence for a correction of the luni-solar precession (ψ) of the order of Δψ = −3 milliarcseconds per year (Williams et al., 1994). When studying the motion of bodies of the planetary system from photographic observations, an inertial stellar reference frame is required. Thus, any imperfection of the luni-solar precession would have repercussions on the determination of the orbit dynamics of the celestial body under investigation.

In the interval covering last three decades optical observations of the Sun and major planets have been produced with an unprecedented intensity. Most of the published world-wide observations of the Sun, Mercury, Venus and Mars made from 1960 to 1994 with transit circles, astrographs and astrolabes are incorporated here for investigation of the residual rotation of the stellar system with respect to the dynamical reference frame.

Minor planet observers generally use the Guide Star Catalog (GSC) as the astrometric reference for the reductions of their CCD frames. Two problems are related with GSC. First, it contains no proper motions, and second, it shows severe systematic errors near the plate edges.

I report on the status of a new reduction of GSC and present a first version of STARNET, a catalogue of 4.3 million positions and proper motions using GSC as second, and a new reduction of the Astrographic Catalogue (AC) as first epoch. In STARNET the average rms-accuracy of the proper motions is 5 milliarcsec/year, that of the present-day positions 0.3″.

Astrometrical properties of the Hubble Space Telescope Guide Star Catalogue (GSC) were recently investigated by the same authors through its comparison with the PPM catalogue. In this paper, a new systematic plate-based magnitude dependent error produced by telescope optics and by the incompleteness of plate reduction model has been found after applying a simplified block adjustment procedure. New subroutine for correction of all systematic errors is developed and GSC can be used now as a dense reference catalogue on about 0.4″ accuracy level.

An automatic reduction process is applied to observations of minor planets obtained at Valencia Observatory, using PPM and GSC catalogues. Systematic corrections of GSC, obtained by the same authors in a previous study, are also considered. The comparison of results is done by the statistical analysis of several parameters. Preliminary conclusions about the improvement obtained using corrected GSC are included.

We prepared the results (in a computer-readable form) of the Belgrade ZT observations made in the period 1949–1985 and finished their new reduction in the FK5 reference frame.

The indirect method of linking the Hipparcos reference frame to the frame defined by extragalactic sources is described. To this end, two independent time series of Earth orientation parameters observed by two different techniques with respect to the two reference frames are used: a) Optical astrometry observations (referred to Hipparcos stars), b) VLBI observations (referred to extragalactic objects). The parallel use of both techniques during the last decade enables to determine the orientation of the two reference frames at a fixed epoch and their mutual slow rotation with precision of at least 1mas and 1mas/year, respectively. In order not to raise confusion, the potentiality of the method is demonstrated on the example based on the star catalogues originally used at the participating observatories, not on any of the existing preliminary versions of the Hipparcos catalog.

This paper describes a proposal for a new space radio astronomy mission for astrometry which uses very-long-baseline interferometry (VLBI) called Astrometry VLBI in Space (AVS). The ultimate goals of AVS are to improve the accuracy of radio astrometry measurements to the microarcsecond level in one epoch of measurements and to improve the accuracy of the transformation between the inertial radio and optical coordinate reference frames. The scientific objectives of the mission cover a few categories of astrometry tasks such as astrometry of the solar system, reference frames ties, tests of general relativity and cosmology (for references see Lowe and Treuhaft, 1994, Russell et al., 1992, Eubanks et al., 1994).

A regular program of observations of the Sun with prismatic astrolabes is under way since 1974 in both, the “Abrahão de Moraes” (φ = −23°0.1′) and “Caussols” (φ = +43°44.9′) observatories. The zenith distance of 2968 transits of the solar limbs observed during the period from 1988 January to 1994 April, in both centers, are analysed. The primary goal of this programme is to determine corrections ΔE and ΔA to the FK5 equinox and equator, as well as to the Earth's orbital constants. Secondly, we present here possible variations of the solar radius, which are of interest because of their astrophysical significance. The results of the analysis of the observations made with these instruments depend on both the prism angles used during the observation and the latitude of the site (Leister 1989, Poppe 1994). With the use of the different reflector prisms, instead of the refracting equilateral, it was possible to observe at differents zenith distances, allowing to sweep, at the brazilian observatory, the apparent orbit of the Sun during a period of about 10 months per year, with two prisms. The instrument of the french observatory works with 11 differents zenith distances, permitting an increase of about 10 months per year in the observational period either.

Positional observations of the Sun have become, in recent years, one of the most important contribution of astrolabes to fundamental astrometry. After pioneer observations at CERGA and Sao Paulo, both in 1974, other astrolabes have been adapted for observations of the Sun in Paris (now at Malatya, Turkey), Santiago de Chile (Chollet & Noël 1993) and San-Fernando (Sánchez et al. 1993, 1995). First experimental campaign of solar observations with an astrolabe installed at Simeiz Observatory (Crimea, Ukraine) has been undertaken in 1986 (Kolesnik 1987). After some instrumental improvements, observations covering about 2.5–3 months were continued in 1987, 1990, 1991. The results are reported here.