The program for establishment of a link between the fundamental system FK5 and the radioastronomical coordinate system is described. The program includes photographic and meridian observations of extragalactic radio/optical sources and intermediate reference stars. Observatories of the USSR, GDR and Yugoslavia are participating in the project.

Some aspects of the realisation of a celestial reference frame using a space astrometry facility are considered. An observational program is described, consisting of observing stars up to magnitude 14, radiostars and bright QSO's, planets, asteroids and of laser signals from the Earth. A scheme of an astrometric facility consisting of two telescopes on board the satellite is proposed. The overview strategy with “inita” is estimated, and the estimates of the accuracy of a single observation (0.″02-0.″05) and of the output catalogues (0.″001-0.″007) are made.

The R.A. system KSV [1] has been used in the USSR Time Service for observations with small transit instruments since 1970. Some methodological errors were made when creating the KSV [2] and the use of the proper motion system, distorted with the position errors of GC and KSV [3], make it necessary to improve the KSV system. Work on the creation of a new system KSV-2 was organized for this purpose. The catalogue included about 0.3 million computer-readable observations with small transit instruments of the USSR Time Service with mean standard deviations σ = 11 sec δ ms.

An estimation of Δδα systematic errors of the FK5 has been made. Systematic differences between the FK5 and FK4 and Δδα corrections to the FK4 derived at Pulkovo have been used as given in the following series: Δδα = A′ sin α + B′ cos α + C′ sin 2α + D′ cos 2α.

The present-day realization of inertial coordinate system by means of traditional optical astrometry is the Fifth Fundamental Catalogue (FK5). The series of observations with new meridian instruments (CAMC, PMC-190, BAMC, HMC) show the significant correlated differences (up to 0.″1) of observational catalogues from the FK5. Moreover, FK5 mean epoch appears to be old (about 50 years ago) and FK5 proper motions would have essential errors owing to the fact that not many new original catalogues (only 25 in RA and 15 in DEC) were used when compiling FK5. It should be noted also that FK5 has a dissimilar accuracy of positions and proper motions of “old” and “new” stars.

The photographic and VLBI positions of 59 quasars are used for determination of the FK4 orientation parameters with respect to the radioastrometrical system (RAS).

The classical Ertel-Struve vertical circle of the Pulkovo Observatory was astablished at the Kislovodsk Mount Station. From 1984 to 1989 we have obtained 566 declinations of the Sun, 230 of Mercury, 413 of Venus, and 207 of Mars. Observations were visual, and the reductions were differential in the FK5 system. Only day-time observations of reference stars were used.

Some information on the construction of the AMC was given at IAU Colloquium 100 in Belgrade. The present paper reports the first results of trial observations of star right ascensions. At present the instrument is being prepared for semiautomatic absolute and relative night-time and day-time observations. The scheme of the instrument, principal units and calculations of their parameters are estimated and an observational programme is proposed.

The MAHIS is the L.A. Sukharev meridian automatic horizontal instrument. It has all the advantages of the Pulkovo HMC and is devoid, or nearly devoid, of its disadvantages. A scheme of the MAHIS includes a flat mirror of D = 300 mm and two immovable horizontal tubes of D = 190 mm, focal length 8000 mm. There is a pavilion for the central unit and 2 roofed pavilions for eyepiece parts of the tubes. The MAHIS is provided with an automatic registering device, a system for meteorological data collection, and is computer controlled. The work on this telescope is being carried out at the Pulkovo, Kazan and Nikolaev observatories. The expected precision of the instrument is ±0.″05 in both coordinates with respect to systematic errors.

Past and present construction of astrometric instrumentation at Pulkovo is discussed.

This work has been made in accordance with the plan of the International Working Group “Double Stars: HIPPARCOS Input Catalogue Consortium.” The precise coordinates for the components of the triple stars from the programs of the Astronomical Observatory (AO) of Leningrad State University (LSU) and the Uccle zone (declination from +30° to +50°) are determined. The work has been made by two methods: (1) special astrometrical observations for about 70 triple stars; some colleagues from the observatories in Pulkovo, Nikolaev, and Goloseevo (USSR), as well as from Belgrade (Yugoslavia) have taken part in this work; (2) statistical treatment for about 200 triple stars, using the data of the catalogues ADS and WDS—all available positional observations of the relative coordinates of the components.

Also, the special observations of the triple stars in Pulkovo and Belgrade have been used, as well as the data from the catalogue by Ch. Worley from 1820 to 1988. The old observations by W. Struve in 1820-1830 play an important role. The data obtained are used for the compilation of the observational programs of the AO of LSU by the HIPPARCOS and HST.

The long series of meridian observations of the Moon can be used for the precise determination of the equinox- and equator-corrections of a star catalogue. Systematic errors of different charts of the lunar marginal zone used for the reduction of the lunar limb observations have no influence on the determination of the secular variations of the zero-points of the fundamental coordinate system.

From meridian observations of the lunar limb made during the interval 1923-1977 in Washington, Greenwich, Cape and Tokyo the following estimate is found for the correction to the right ascension system of the FK4 catalogue: which is in disagreement with the values used for the compilation of the FK5 catalogue.

A notion of density function of systematic errors of astrometric catalogues distributed on the unit sphere as a simple layer is introduced. Components of the catalogue errors in any direction are determined as partial derivatives of layer potential in the same direction. For example, the density function of FK4 errors is computed as an expansion of spherical harmonics.

A programme of Talcott pairs has been compiled at Pulkovo in order to tie bright radio sources to the FK4 system. The programme includes FK4 stars and radio stars to mag 9 (Radio stars for astrometry, Abh. Hamb. Sternw., 1982). About 150 observations of 18 pairs were made during 1985-1989. The correction to the declination of a radio pair has been found as a difference between the “reference” latitude and the latitude derived from the observation of the radio pair. The value of the “reference” latitude was determined from observations of ordinary Talcott pairs. The standard deviation does not exceed 0.09 arcsec. A radio star can be tied to the FK4 with an accuracy not worse than 0.1 arcsec.

A circle-reading device of the meridian circle with a constant link to the vertical line and the method of using the device are proposed. The accuracy of a circle reading is expected to be 0.″02.

Krejnin and Muni's method (1973) enablesenables one to derive absolute declinations of stars in a narrow equatorial zone |δ| < 10′ from observations near the Earth's equator |ϕ| < 10′. Some systematic effects, including the errors of the value of the micrometer screw for two equatorial instruments (or the scale error if one of the instruments is a PZT), might be determined if a global reduction is used for the original observations from the equator and from those of an astrolabe at latitude |ϕ| ∼ 20° to 23°. Astrolabes—especially photoelectric ones (Hu 1988) are considered to be the most efficient for determination of absolute declinations of stars and absolute latitudes of the instruments in Tolchel'nikova-Murri (1985).

In Izv. GAO No. 206 the method will be published as well as the criterion for estimating the efficiency of different programs, which is required to improve planning in astrometry.

Astrometry is a branch of science which develops methods for the quantitative descriptions of places and time instants of astronomical events on the basis of observations of celestial bodies. For this purpose a theoretical coordinate system is introduced (e.g. equatorial a, α, δ). The aim of astrometry is to apply this system to the observed reference objects (stars, planets etc.) so that their coordinates α(t), δ(t) can be calculated according to the relations α(t) =f1(Pk, t-to) and δ(t) =f2(Pk, t-to) where Pk are parameters, to is the conventional time instant and t is the current time. In order to understand the term inertial coordinate system assume that the coordinates α(t), δ(ti), i=1,2,…n are used for plotting the coordinate origins. If these coincide then the system is conventional-fixed and therefore inertial. Thus, the inertial coordinate system in astrometry is a conventional-fixed reference frame reproduced with the use of celestial bodies whose law of motion is known with sufficient accuracy.

The work on photographic astrometry in Bolivia begun in 1983, when an expeditional astrograph was put there (D=23 cm, F=230 cm, working field of 4°x4°). The main purpose of the observations with this telescope is extending the inertial coordinate system to different classes of stars. For this purpose we have carried out and are currently carrying out the following observational astrometric programs:

1. The preliminary catalog of precise positions of 200 000 southern stars to mag 11 with declination in the range from 0° to −90° was created. Each star was observed four times with displacement of the photoplate by 2° in δ and by 8 secδ in α. The exposure time is from 4 to 12 minutes. The accuracy of the preliminary catalog is ±0.247 arcsec and the final one is ±0.15 arcsec.

2. We are not far from finishing the work on a catalog of 1900 bright stars (−90° < δ; 0°, m ≥ 6.05). Each star was observed not less than 2 times for different positions of a special filter which diminished the light of bright stars.

3. The work on astrometry of equatorial stars to mag 12 in the range of declination from −20° to +20° was begun. From the beginning of the observations on 1 May 1987, 1150 astroplates were obtained. On the whole it is necessary to obtain 3840 photoplates. All three programs use the reference catalog SRS, compiled by Pulkovo and Washington astronomers, in the FK5 system.

The methods of the rational leveling of the results of the observations of the angular distances between quasars by VLBI method were suggested. The application of these methods on the model of reserved net on sphere shows a descrease of the standard error of the quasers' coordinates after leveling by a factor of three times. We employed 10 elementary configurations constructed on 20 chosen quasars in the Northern hemisphere.

The concept of reference system, reference frame, coordinate system and celestial sphere in a relativistic framework are given. The problems on the choice of celestial coordinate systems and the definition of the light deflection are discussed. Our suggestions are listed in Sec. 5.