Observations of lunar occultations made between 1800 and 1955 are analyzed in order to determine the fluctuations of the Earth's rotation before the International Atomic Time was established. The reduction is made on the basis of the DE200/LE200 ephemeris and the FK5 system. The analysis of the fluctuations shows a main period of 11.7 lunations caused by the error in the stars places.

A new analysis of the timing data acquired on the fast pulsar PSR1937+214 is presented. Parameters are evaluated with various models based on two ephemerides, two atomic time scales and two TT—TB time transformations. Comparisons are carried out with results from other programs. We provide evidence that systematic errors induced by the model adopted are 5 to 10 times larger than the formal uncertainties calculated by the fitting procedure. Great care must thus be taken when using results from different millisecond pulsars timing programs for accurate astrometric purposes.

We propose a quasi-inertial four-dimensional reference frame in the solar system asymptotically-flat spacetime on the basis of pulsar astrometry techniques. The Pulsar Reference Frame (PRF) consists of the Pulsar Time Scale (PT) and the the Pulsar Reference System (PRS) on the sky.

POINTS is a space-based optical astrometric interferometer, capable of measuring the angular separation of two stars about 90° apart with 5-microarcsec (μas) nominal accuracy. During the intended ten-year mission, a repeated survey of a few hundred targets over the whole sky, including a few bright quasars, will establish a “rigid” reference grid with 0.5 μas position uncertainties. At that level, the grid will be free of regional biases and tied to the extra-Galactic frame that is our present best candidate for an inertial frame. POINTS will also determine parallaxes and annual proper motions at about the same level. Further, the planetary ephemeris frame will be tied through stellar aberration to the grid at about 300 μas. Additional targets of interest, to a limiting magnitude of greater than 20, will be observed relative to the grid, yielding determinations with uncertainties depending on the observing schedule. Measurement at the microarcsec/year level of the apparent relative velocities of quasars that are widely separated on the sky will severely test the assumption of cosmological quasar distances and may also constrain models of the early Universe.

The framework of general relativity theory (GRT) is applied to the problem of reduction of high precision astrometric observations of the order of one microarcsecond. The equations of geometric optics for the non-stationary gravitational field of the Solar system have been deduced. Integration of the equations of geometric optics results in the isotropic geodesic line connecting the source of emission (a star, a quasar) and an observer. This permits to calculate the effects of relativistic aberration of light due to monopole and quadrupole components of the gravitational field of the Sun and the planets taking into account their motions and rotation. Transformations between the reference systems are used to calculate the light aberration occurring when passing from the satellite system to the geocentric system and from the geocentric system to the baryecntric system. The baryecntric components of the observed position vector reduced to the flat space-time are corrected, if necessary, for parallax and proper motion of a celestial object using the classical techniques of Euclidean geometry.

The astrometric optical interferometer on Mt. Wilson is providing a new source of astrophysical and astrometric data on bright stars. The instrument, with 12-meter baselines, has been in operation since late 1986. The interferometer is capable of wide-angle astrometry, that is, the determination of very precise stellar positions within a reference frame defined by bright stars spread across a large area (of order one steradian) of the sky. This paper addresses the question of whether such an instrument can be used to establish a fundamental system — that is, one tied to the Earth in some well-defined way. Some astrometric data from this instrument are presented to illustrate the difficulties involved. Proposed means of addressing these problems in future instruments are discussed.

Inertial reference frames spanning approximately 10°-30° square on the sky and capable of locating objects to few-hundred microarcsecond accuracies are useful for a broad class of astrometric measurements. Deep space tracking and general relativistic angular deflection experiments are examples of astrometric measurements which can profitably reference the positions and/or motions of objects to a field of radio sources in a local frame. A method for defining local inertial reference frames has been developed based on Very Long Baseline Interferometry (VLBI) measurements of extragalactic radio sources. By observing the radio emission from the object to be located in the frame, as well as that from about five radio sources which define the frame, dominant systematic astrometric errors can be minimized through parameter estimation. The entire reference frame measurement is of the order of 30 minutes including all the sources in a frame. The limiting error for single-epoch position determination in a local frame is the unknown structure of both target and reference objects. Structure can cause systematic milliarcsecond-level errors. The limiting error for epoch-to-epoch differential position measurements is tropospheric fluctuations, assuming that the radio source structures do not change from one epoch to the next. Preliminary results of an epoch-to-epoch measurement of relativistic gravitational deflection by Jupiter, in which the total deflection was about 600 microarcseconds, suggest that the local reference frame is stable at the 240-microarcsecond level over twelve days. Data have been taken at longer time intervals to determine the annual stability of the frames. At the time of preparation of these proceedings, those data have not yet been analyzed.

Assessment of the impact of recent improvements in Deep Space Network (DSN) instrumentation, as well as of joint data analyses, provide a prognosis for the accuracy level to be expected in future realizations of an inertial radio reference frame. Intercontinental dual-frequency radio interferometric measurements during 68 sessions (including two recent sessions employing Mark III instrumentation) from 1978 to 1989 using NASA's DSN stations in California, Spain, and Australia give 8900 pairs of delay and delay rate observations. Analysis yields a catalog of positions of 200 extragalactic radio sources north of —45° declination. The resulting source position formal uncertainty distributions peak below 1 milliarcsecond, with three fourths being smaller than 2 mas. Comparison with independent measurements shows some evidence for systematic errors at the milliarcsecond level.

Over 350 000 dual frequency Mark III VLBI observations from several geodetic and astrometric observing programs have been used to realize an inertial reference frame through the positions of 325 compact extragalactic radio sources uniformly distributed over the sky with standard errors typically under 1 milliarcsecond (mas). Internal and external tests indicate that the reference frame defined by the relative positions of these radio sources should be accurate and stable at the 1-2 mas level. Because the conventional precession and nutation models are adjusted in the estimation of the source positions, the positions and relative angles are not degraded over the interval of observations or at epochs away from the reference epoch.

In 1987 we began a 5-year program to establish a reference system of at least 400 extragalactic sources which are compact and flat spectrum in the radio and which also display optical emission. This reference frame is to be global with about one source per 100 sq deg (Johnston, et al., 1988). The program incorporates some data which had been obtained previously for other purposes. Altogether, in one or another aspect of the program, 489 sources have been considered so far.

In an attempt of a realization of the radio reference frame a compilation catalogue of positions is derived from independent observation catalogues of extragalactic objects the coordinates of which had been determined by means of Very Long Baseline Interferometry. The compilation catalogue comprises 209 objects which are divided into a core consisting of 50 objects having mean positional accuracies of 0.5 milliseconds of arc (mas) and an extension with positional accuracies better than 2 mas. Comparison of this catalogue with an independent compilation catalogue led to confidence limits at the 1 mas level. - The compilation catalogue is supposed to represent a static reference frame of fixed extragalactic points. As the epochs of the contributing observations span nearly 10 years it was tried to interpret the apparent motion of the fixed points recognizable in the observation catalogues as an effect of luni-solar precession. The pilot study points at a reduction of the conventional value of about 2 mas per year.

The radiointerferometrical network “QUASAR”, composed of six dedicated VLBI stations linked via a geostationary satellite channel with the Center of Operations, is under construction in the Soviet Union. It is proposed to construct “QUASAR”-stations abroad, in China, Bulgaria and India. A short review of basic scientific and technical features of the project is given in this paper.

An attempt of the realisation of the radio reference frame was undertaken. For this purpose a new approach to the construction of a compiled catalogue from observation catalogues of extragalactic radio sources was used. As a numerical example a compilation catalogue of primary radio sources based on the three individual catalogues (GSFC, JPL, and NGS) was constructed. Comparison of this catalogue with recent compiled catalogues led to the conclusion that their r.m.s. differences were about 1 mas.

The Hipparcos satellite was successfully launched by Ariane 4 flight 33, from French Guyana, on 8 August 1989. However, the apogee boost motor, designed to place the satellite into its intended geostationary orbit, failed to function, and the satellite was left in a highly elliptical transfer orbit. After a perigee raising manoeuvre, using the hydrazine propulsion system, the orbital parameters are approximately: apogee 35900 km, perigee 526 km, eccentricity 0.72, inclination 6.8°, orbital period 10h 40m.

Operational procedures have had to be revised during the first weeks of satellite operations, due to the non-nominal orbit. Commissioning and calibrations have taken longer than foreseen because of the operational difficulties and observational overheads, but routine observations commenced on 27 November. The essential differences between the nominal and the present missions are that observations can only be conducted for about 30-40 per cent of real time in the present mission, and the fact that the actual lifetime of the mission is still rather uncertain, and possibly somewhat shorter than anticipated. This paper gives a summary of the present satellite status, measurement results obtained so far, and in indication of the scientific goals that may be achievable with the present mission.

The Hipparcos mission had to be revised because the satellite did not reach the circular geostationary orbit. Observations from the elliptical transfer orbit will be degraded in the sense that good quality observations are expected, but only during less than 50 percent of the total time, and the mission duration will probably be less than 2.5 years. If the mission lasts 12 months a Tycho catalogue is expected containing at least 200 000 stars with typical accuracies of 0.10 arcsec for positions and 0.10 mag for B and V magnitudes.

The Sphere determined from the observations carried out by the ESA satellite Hipparcos will have a residual rotation with respect to distant extragalactic sources. The alignement of the two systems of reference will be made at the mission completion. The procedure to be applied is discussed along with its expected accuracy.

The strategy for linking the HIPPARCOS Catalogue to extragalactic objects originally devised for the geostationary orbit, is no longer adequate for the elliptical orbit to which the satellite is now confined, but useful improvements in the rotation matrix may be obtained by incorporating proper motions from certain other catalogues and from ground based absolute proper motions.

The observation of minor planets by Hipparcos offers the opportunity to obtain high precision positions for some minor planets. About fifty minor planets are on the programme. Their ephemerides had to be improved in order to reach a precision of 1 arsec and occultations by the Earth and the Moon had to be predicted.

From the position of a minor planet on reference great circles at different times better values of the initial position and velocity will be deduced but the reduction of the observations of the minor planets have to take into account the displacement of the photocentre relative to the centre which is due to the shape, the phase effect and the scattering properties of the surface. For some very small planets considered as star like this diplacement will be small and the precise positions obtained will allow to position the dynamical reference system relative to the Hipparcos system. For the bigger minor planets the observations by Hipparcos may give informations on the shape and scattering properties of the surface.

The question concerning the influence of the consideration of asteroid mutual perturbations on the calculated positions of 20 selected minor planets as well as on the fundamental catalogues systematic corrections is discussed.

The Hubble Space Telescope, a large optical instrument having an aperture of 2.4 meters and a length of 8.8 meters has been developed by the U.S. National Aeronautics and Space Administration in cooperation with the European Space Agency. The Space Shuttle will be used to place the telescope in orbit. The primary astrometric instrument will be one of the three Fine Guidance Sensors which have the capability of measuring the position of one object with respect to another to an accuracy of ±0.″002. To facilitate use of the Hubble Space Telescope, observers will be provided with the Astrometric Data Reduction Software package. The variety of astrometric problems and the several modes of operation are mentioned as well as the cooperative program with the European astrometric satellite project HIPPARCOS.