Three windows towards the Galactic bulge, located at (l, b) = (1.6,-2.8), (5.2,-3.5), and (4.2,-4.8), were identified from deep astrometric plates. Deep BVRI CCD photometry of these fields reaching V = 18.5 was obtained at the ESO Dutch telescope. The color-color and color magnitude diagrams of these fields allow us to select different bulge tracers, including blue horizontal branch stars, red clump giants, and RGB tip stars, as well as disk main-sequence stars. A first catalogue of some 2000 red clump giants was constructed, and measurement of their proper motions in under way. We are deriving absolute proper motions in the Hipparcos system using a two step procedure. First, plates from the Cordoba archive of the Astrographic Catalogue (epochs 1903-1913) are used in combination with recent plates from theESO GPO telescope to tie a set of intermediate magnitude field stars in the Hipparcos system of positions and proper motions. Second, deeper intermediate epoch (1958) and final epoch CCD images (epoch 1996.5) are used to determine the absolute proper motions of the fainter bulge stars. The average uncertainty in the absolute proper motion of a single giant, based on a preliminary sample of few tens of stars, is 0”.006 yr-1. Thisrelatively small error, in combination with the sample size, would allow usto study the shape of the velocity ellipsoid for the different Galactic components present in the inner regions as function of Galactocentric distance, and also to measure an accurate distance to the Galactic center and themass of the bulge. In addition, we identified about 20 faint objects which are potential members of the Sgr dwarf, based on their location in the color-magnitude diagrams, and are also measuring their proper motions.

Other windows with measured proper motions are located along the Galactic minor axis, including Baade’s window at l, b = (1, −4) (Spaenhauer, A., Jones, B. F., &: Whitford, A. E., 1992, A3, 103, 297), and the Plaut field at l, b = (0, - 8 ) (Mendez, R. et al., 1997, in 4th ESO/CTIO Workshop on “The Galactic Center”, in press). Our proper motions are complementary, probing the kinematics off the Galactic minor axis, where the signature of rotation should be evident. Thefields chosen for this study overlap fields that the MACHO project is following since 1993. Their variable star database would provide with excellent tracers of different populations. The 3-D motions of these different Galactic components would be measured for the first time. These include RR Lyrae tracers of the inner metal-poor halo, Miras, LPVs and delta Scuti stars tracers of the metal-rich bulge, and eclipsing binaries tracers of the Galactic disk (Minniti, D., et al., 1996, in IAP Coll. on “Variable Stars and the Astrophysical Returns of Microlensing Surveys”, ed. R. Ferlet, p. 257; Alcock, C, et al., 1997, astro-ph/9707311).

Using Hipparcos parallaxes and proper motions together with radial velocity data and individual ages estimated from isochones, the velocity ellipsoid has been determined as a function of age. On the basis of the available kinematic data two different samples were considered: a first one (7789 stars) for which only tangential velocities were calculated and a second one containing 3104 stars with available U, V and W velocity components and total velocities ≤ 65 km.s-1.

The main conclusions are: -Mixing is not complete at about 0.8-1 Gyr. -The shape of the velocity ellipsoid changes with time getting rounder from σu/σv/σ-w = 1/0.63/0.42 ± 0.04 at about 1 Gyr to1/0.7/0.62 ±0.04 at 4-5 Gyr. -The age-velocity-dispersion relation (from the sample with kinematical selection) rises to a maximum, thereafter remaining roughly constant; there is no dynamically significant evolution of the disk after about 4-5 Gyr. -Among the stars with solar metallicities and log(age) > 9.8 two groups are identified: one has typical thin disk characteristics, the other is older than 10 Gyr and lags the LSR at about 40 km.s-1 . -The variation of the tangential velocity with age(without selection on the tangential velocity) shows a discontinuity at about 10 Gyr, which may be attributed to stars typically of the thick disk populations for ages > 10 Gyr.

The analysis of stellar kinematics of the local system of young stars from Hipparcos data offers an excellent opportunity to revise models accounting for the origin of the observed deviations from circular motions, and to give new hints on the history of star formation and the dynamical processes involved in the evolution of our Galaxy. We have complemented the astrometric Hipparcos data with a careful compilation of available radial velocitiesand Strömgren photometry, information which enables us to obtain reliable space velocities and individual ages for a large number of stars. Thevelocity field has been studied by means of the classical first order approach, interpreting the trends observed in the Oort’s contants in terms of the local expansion of the Gould’s Belt (stars with ages less than 30-50 Myr, at heliocentric distances ρ < 400 — 600 pc). The fundamental geometric parameters of such structure have been determined. Several statistical methods have been applied to identify moving groups in the 4-dimensional space of (U,V,W, age). Our results definitively confirm their existence with mean velocity dispersions of the order of 5-7 km/s. The obtained results will be published in Astronomy & Astrophysics.

As far as the standard picture of galactic evolution and structure is concerned, early-type stars belong to disk populations. However, the existence of a population of early-type stars with nearsolar metallicities and main-sequence surface gravities, together with high velocities and/or standing at large distances away from the plane, is a well known anomaly to this picture.

In order to provide an insight into the different populations of early-type stars, a sample of high-velocity stars has been gathered from the Hipparcos catalogue. We have selected the stars according to B — V < 0.4,and a tangential velocity larger than 80 km/s. The sample is composed of 101 stars.

The high-velocity early-type stars in the solar neighbourhood are a mixture of different stellar populations: blue horizontal halo stars passing near the sun and high-velocity disk stars. We have decomposed into gaussiansthe V and W-distributions of our sample, to isolate the disk from the halo populations. The U-distribution of high-velocity disk stars is clearly asymmetric. The selection bias over the tangential velocities creates a depletion of low U values. Moreover, two thirds of the sample have U < 0 (toward the galactic anticentre, l = 180°). This cannot be explained by the distribution tail of disk stars or stochastic events such as runaway stars, which privilege no direction.

We have combined radial velocity measurements with Hipparcos proper motion and parallax data for 552 nearby stars of a selected sample of 1208 stars to find which stars may have passed (or will pass) close enough to the Sun to perturb the cometary Oort cloud. We find that the rate of close stellar approaches within a distance D (in pc) from the Sun is N = 4.3 D2.02 Myr-1 in excellent agreement with the predicted functional form of simple stellar dynamics models though the number of stars is less than predicted. However, we consider this a lower limit because there isevidence for observational incompleteness in our data set. One star (Gliese710) is found with a future closest approach of less than 0.4 pc, and several stars come within about 1 pc during about a ±10 Myr interval. We have performed dynamical simulations which show that none of the passing stars in this sample can perturb the Oort cloud sufficiently to create a substantial increase in the long-period comet flux at the Earth’s orbit.

The Hipparcos data are providing a dramatic increase, qualitatively and quantitatively, of the basic available distance information. For example, the numbers of stars for which a relative accuracy better than 10 % is available from Hipparcos data and from ground-based data are respectively 22 396and about 1000. Moreover, the range of spectral types and luminosity classes for which precise parallaxes are available is considerably enlarged, including many stars in open clusters and a small number of Cepheids and RR Lyrae. The bottom of the main sequence is populated down to absolute Hpmagnitude 14, including a number of subdwarf stars essential to derive globular clusters distances and ages (Pont et al, 1997a).

Finally, the Hipparcos data show how difficult are the calibration of photometric distances and the transformation of relative trigonometric parallaxes to absolute parallaxes. This is illustrated by the comparison of distances given in the last edition of the Catalogue of Nearby Stars (CNS3, Gliese & Jahreiß 1991), which is the best available compilation of stars said to be closer than 25 pc from ground-based data. About a third of them are found by Hipparcos to be (much) further than this limit (Perryman et ai, 1995). A second example is given by the study of (Binney et al, 1997).

The Hipparcos parallaxes and proper motions together provide a consistent picture of the Hyades distance, structure and dynamics. They yield a cluster convergent point motion consistent with the individual trigonometric parallaxes, and together explain the large distance modulus derived from the most recent ground-based proper motion investigations as originating from differences in the space velocity and small systematic effects in the ground-based proper motions. Conversely, the smaller distance modulus generally derived from a variety of ground-based trigonometric parallax programmes originates from systematically larger values of the ground-based parallaxes. Although recent distance determinations to individual objects in the cluster, most notably the results of Torres et al. (1997a, b, c), are in excellent agreement with the Hipparcos trigonometric parallaxes, extrapolation to a mean cluster distance is again affected by systematic effects in the ground-based proper motion studies.

The combination of the Hipparcos astrometry with radial velocity measurements from ground-based programmes provides three-dimensional velocities allowing candidate membership selection to be based on 3-dimensional positional and kinematical criteria. A number of new cluster members have been found within 20 pc of the cluster centre. No evidence for systematic internal velocity structure is found; rather, the results are fully consistent with auniform cluster space motion with an internal velocity dispersion of about 0.3 km s-1. Spatial distribution, mass segregation, and binary distributions are consistent with N-body simulations.

13 clusters have been analysed (Coma Ber, the Pleiades, IC 2602, IC 2391, Praesepe, Alpha Per, NGC 6475, Blanco 1, Stock 2, NGC 6633, IC 4756, NGC 2516 and NGC 3532). Cluster members have been selected using Hipparcos proper motions and parallaxes. The mean cluster parallaxes have been determined using Hipparcos intermediate data (abscissae on the Reference Great Circles) taking into account star to star correlations. Both selection of members and computation of cluster mean parallaxes are described in Mermilliod et al., 1997 and Robichon et al., 1997.

For each cluster, a selection of fainter stars in the ‘Base des Amas’ (Mermilliod, 1988), based on photometric data as well as radialvelocities and proper motions when available, has been done. From this selection and the computed mean parallaxes, the location of the 13 cluster sequences in the observational HR diagram is derived very accurately, at least for the closest clusters for which good photometry and small errors on the mean parallaxes are available. Relative positions of cluster sequences cannot only be explained by differences between cluster metallicities. For example, the Pleiades sequence is 0.5 magnitude bellow the Praesepe sequence whereas the adopted metallicities (obtained spectroscopically i.e. whithout any assumption on the helium content) are respectively -0.034 and +0.10.

The present determination of the absolute magnitude .Mv(RR) of RR Lyrae stars is twofold, relying upon Hipparcos proper motions and trigonometric parallaxes separately. First, applying the statistical parallax method to the proper motions, we find < Mv(RR)>= 0.69 ± 0.10 for 99 halo RR Lyraes with <[Fe/H]> = -1.58. Second, applying the Lutz-Kelker correction to the RR Lyrae HIP95497 with the most accurately measured parallax, we obtain Mv(RR) = 0.57-0.74 at [Fe/H]=-1.6. Furthermore, allowing full use of low accuracy and negative parallaxes as well for 125 RR Lyraes with - 2.49≤[Fe/H]≤0.07, the maximum likelihood estimation yields the relation, Mv(RR)= (0.59±0.37)+(0.20±0.63)([Fe/H]+1.60), which formally agrees with the recent preferred relation. The same estimation yields again My (RR) = 0.65 ± 0.33 for the 99 halo RR Lyraes. Although the formal errors in the latter two estimates are rather large, all of the four results suggest the fainter absolute magnitude, My(RR)=0.6-0.7 at [Fe/H]=-1.6. The present results still provide the lower limit on the age of the universe which is inconsistent with a fiat, matter-dominated universe and current estimates of the Hubble constant.

It has been suspected (Szabados 1997) that the apparent orbit of previously unresolved binaries involving a Cepheid component falsifies the Hipparcos-parallaxes deduced for these Cepheids. Since the incidence of binaries among Cepheids exceeds 50 per cent, this adverse effect on the parallax determination is a major error source.

An obvious remedy is to repeat the determination of the five astrometric parameters for the Cepheids, members in binary systems, taking into accountthe orbital elements known from spectroscopic studies. This will result in more precise parallaxes and cosmic distance scale. Moreover, these Cepheidsare ideal targets for the future microarcsec astrometric missions.

This study was partly supported by OTKA-grants T 014852 and T 022946.

The Hipparcos mission demonstrated the efficiency of space astrometry (in terms of number of objects, accuracy, and uniformity of results) and the fact that a relatively small instrument can have a very large scientific potential in the area of astrometry. However, Hipparcos could probe less than 0.1 per cent of the volume of the Galaxy by direct distance measurements. Using a larger instrument and more efficient detectors, it is now technically feasible to increase the efficiency of a space astrometry mission by several orders of magnitude, thus encompassing a large part of the Galaxy within its horizon for accurate determination of parallaxes and transverse velocities. Such a mission will have immediate and profound impact in the areas of the physics and evolution of individual stars and of the Galaxy as a whole.

DIVA (Double Interferometer for Visual Astrometry) is a Fizeau interferometer on a small satellite. It will perform astrometric and photometric observations of at least 4 million stars. A launch in 2002 and a minimum mission length of 24 months are aimed at. A detailed description of the experiment can be obtained from the DIVA homepage at http://www.aip.de:8080/dso/diva. An overview is given by Röser et al., 1997.

The limiting magnitude of DIVA is about V = 15 for spectral types earlier than M0, but drops to about V = 17.5 for stars later than M5. Table 1 gives a short overview on DIVA’s performance. DIVA will carry out a skysurvey complete to V = 12.5. For the first time this survey will comprise precise photometry in at least 8 bands in the wavelength range from 400 to 1000 nm.

DIVA will improve parallaxes by a factor of 3 compared to Hipparcos; proper motions by at least a factor of 2 and, in combination with the Hipparcos observations, by a factor of 10 for Hipparcos stars. At least 30 times asmany stars as Hipparcos will be observed, and doing this DIVA will fill the gap in observations between Hipparcos and GAIA. DIVA’s combined astrometric and photometric measurements of high precision will have important impacts on astronomy and astrophysics in the next decade.

Dispersed interferometer fringes as a new type of signal are currently under discussion for several space interferometry mission projects (GAIA, FAME, DIVA). The dispersed fringes are directly recorded by CCD mosaics whichare operated in drift-scan mode on board a continuously rotating satellite. They allow combined astrometric (perpendicular to the fringes, i.e. in the direction of the drift scan) and spectro-photometric measurements (in the direction of the dispersion of the fringes, i.e. perpendicular to the drift scan). We have simulated polychromatic dispersed fringe patterns of stars with known spectra. The specific instrument parameters used for the simulation are those envisaged for the DIVA instrument, with realistic assumptions on various noise contributions. The results can be scaled to GAIA and FAME instrument parameters. We have investigated simple algorithms which can be implemented in an on-board data reduction and used for the real-time attitude determination with bright and single stars. On the preliminary basis of 100 simulations per star of a given magnitude and spectral type, we have obtained first estimates of the achievable accuracy of the astrometric measurements and of the location of the photometric zero point λo.

As a follow-up to the most successful Hipparcos mission we have suggested a mini-satellite. The instrument DIVA is a pair of two small Fizeau interferometers with a fixed angular separation scanning the entire sky in a Hipparcos-type fashion. Despite its small apertures the very efficient opticaldesign and the use of CCD detectors will allow a significant increase in performance compared to the Hipparcos mission. Introducing objective prisms in front of the beam combiners will generate dispersed fringes which will allow a more accurate astrometry and much more detailed spectro-photometric information than filtered, undispered fringes. A complete optical system was designed which is suitable to achieve the desired goal and may be expanded to the size of GAIA. Likewise, several options exists to improve the capabilities and the depth of the DIVA mission.

The Space Astrometry and Photometry Project Struve is being designed at the Pulkovo Observatory in cooperation with other Russian space institutes. It is expected to be launched before 2010 with a duration of mission of at least 36 months.

The main objectives are to extend at milliarcsecond accuracy the Hipparcos satellite reference system to fainter objects including quasars and to get a second epoch for Hipparcos stars. The project Struve, along with the recently suggested project DIVA, will fill the intermediate place between Hipparcos and microarcsecond astrometry. Unlike DIVA, we propose far more extensive astrometric and photometric surveys.

We expect an Output Catalogue of 20 million stars (density of about 500 stars per square degree). A sky survey will be complete down to V = 14 (about 15 million stars), and selected objects down to V = 19.5 will be observed within a special program. The proper motions of the Hipparcos stars will be determined with an accuracy of about 0.1 mas/yr. The mean accuracy of star positions in the output catalogue is expected to be 0.6 mas which could be achieved by proper design of the satellite (symmetry, smooth rotation etc.), optics and the micrometer. A properly designed micrometer (with CCD arrays, special processors for image processing and the compression of the data flux to the ground station) will give the possibility of observing all objects of the sky down to a definite limiting magnitude.

In connection with the creation of the scientific grounds of the Russian Project named “Struve” Space Astrometric System, the main principles of classification and identification of any celestial moving object observed with this System were formulated and basic algorithms were elaborated. These algorithms are invariable for a short or long observational arc, for a known or unknown heavenly body and for an artificial or natural celestial object.

For classification of observed celestial bodies the angular velocities of motion are used. These values are calculated from the statistical treatment of the 5 nearest spherical celestial body positions with a time intervalbetween them near 7 seconds.

For an identification of well known sky objects ordinary procedure may be used by means of traditional ephemeris calculations. For an identification of unknown natural celestial bodies a special algorithm was developed. Ituses a calculated angular velocity of moving object from one scan to another close scan of observations. Then, having several accurate positions of fixed object during 5-10 hours per day, we can determine an initial elliptical object’s orbit by the Apparent Motion Parameters Method created at Pulkovo Observatory. It deals with a position of object, its angular velocity and acceleration, position’s angle and a curvature of trajectory on a short observational arc. These last four quantities are named the Apparent Motion Parameters.

In spite of a preliminary character of the AMP-method orbits we can identify an observed object through a large interval of time, for example over 2-4 months after orbit determination. In this problem the observed and calculated angular velocities of the object’s motion are very useful.They are new and important ephemeris parameters at the epoch of Space Telescope astrometric positional observations. Examples of simulations of orbit determinations will be presented.

Seventeen years ago, the Hipparcos project was accepted within ESA’s scientific programme, as the first space mission specifically devoted to the accurate measurement of stellar positions. As the result of an unprecedented and productive collaboration between ESA and the European astronomical community, the Hipparcos project has reached a decisive milestone.

The Hipparcos and Tycho Catalogues were completed in August 1996, and made widely available in June 1997, when they were published by ESA as the 17-volume publication ESA SP-1136. Volume 17 includes all data files, and documentation, in the form of six ASCII CD-ROMs. The catalogues can be obtained from the ESA Publications Division, with full details (and an order form)given on the Hipparcos www page, http://astro.estec.esa.nl/Hipparcos. The full catalogues are also available via ftp from the CDS, Strasbourg, where the catalogue data are also linked to the cross-identifications in the SIMBAD data base, and through the VizieR relational data base queries.

The Hipparcos www page also supports remote interrogation of the catalogues, either by object (HIP or TYC) identifier, or via a search range in right ascension and declination. Active links allow the user to follow selected HIP or TYC identifiers, entering the associated annexes (Variability Annexes, and the Double and Multiple Systems Annex) or inspecting the light-curve (in PDF form) when provided. A more comprehensive search and interrogation facility will become available at the end of 1997 in the form of a CD-ROM (Celestia 2000) with access and display software prepared for IBM PC and compatibles running under Windows 3.1, Windows 95, or Windows NT.