NTT and the quality. - of the food in the La Sllla cafeteria. Entering the Ws,we started to fear that the driving core of tl-ta team was softening with increasing age, that motC vations were on the low side, that new arrivals did not integrate in the possibly obsolete working schemes in short that we might not be up to the new challenges that ESO is facing. I On a hot Saturday in July. on the Max Planck field in Garching, an €SO team wlth a shaken self-confidence and an average age dangerously approach- ing 40, entered the ASTRO CUP, a one- day mmpdtion wlth the teams of our neighbours and friends of the Max Planck Institutes fiir Astrophysik and fiir Extraterrestrische Physik and of the Obsewatory of the University of Munich. At the end of the day, after four strenuous games where w scored 5 I I L- 4-11 r WIW JW-~~~~~Lvrllpmm - "--", Ulrl rC1 Iw nmua, ". goals and suffered 2, we stmd as Gouiffea, a visitof to ESO, E. ZoN (a friend from NW,A+ W~ltander,E Koeh, J. -tie; baHered but winners with cup b0tlom: L Noethe, Zigmnn, BB,belabre, & D'm,#. mttn*,#. 6asbiItr (ofher team at our feet and glases of ext3ellentmmbm not includ& in this pictwe: B. Buzzoni. D. Chittim, A, van Dijs~~Idonk,G. fishw, P. Fmnpis, 8. Jevgensen, P. Mellm T. and R. Wannets). Bavarian beer in our hands (both cour- Oostedm tesy of the sponsor CONVW). We might well lose the Cup next year to one of our excellent contenders, but we are satis- have energies to spend when needed. tivities of the Organiratlon, but it does fied to have proved this time that we are We are too realistic to claim that this not hurt to secretly play with this feeling. not at our wit's end and that we still victory is a good omen for other ac- LongliveESO! S. D'ODORICO, ESO Astrometry with ESO Telescopes A Contribution to the Construction of the New Extragalactic Reference Frame Chr. DE VEGT, Hamburg Observatory, Germany Astronomical research is strongly de- be presented to the IAU during the 1994 history, but only in the last decade the pendent on the availability of a unique General Assembly in The Hague. practical realization of this concept be- all-sky reference frame though most as- came feasible through the mature tech- trophyslcists do not explicitly take nique of VLBl radio astrometry. notice of this complex astrometric 1. Main Properties of the New Using a global net of suitably distri- problem. Reference Frame buted radio telescopes, positions of However, the necessity of very pre- these primary radio sources can now be cise pointing of new generation large Contrary to the present fundamental determined in a routine way to milli- telescopes from ground or space and system which is based on the positions arcsecond (mas) precision and an abso- the unambiguous identification of very and proper motions of bright stars - the lute global reference frame can be es- faint objects in all spectral regions ac- basic FK5 contains 1535 bright stars -, tablished and maintained for the future. cessible from ground and space, in par- the future extragalactic system will be At the same time the high angular ticular in the radio and infrared region, based primarily on the positions of a resolution of VLBl provides comprehen- has sensltired the astronomical com- carefully selected small number of corn- sive information on source structure and munity to this problem. pact extragalactic radio sources;almost their temporal changes with sub-mas During the IAU General Assembly in alt of these sources will display optical resolution. Buenos Aires a resolution by the Work- counterparts, mainly quasars and BL A second group of objects is of equal ing Group on Reference Systems has kc's but also some compact galaxies. importance for solving this problem; been adopted (IAU 1992) which de- This choice is based on the generally namely selected radio stars, the cm scribes the properties of a new, inertial, agreed assumption of cosmic distances radio emission of which has to be strong extragalactic reference frame and a new of these objects with the consequence and steady enough to be measured with intercommission working group has of negligible proper motions and there- mas precision by VLBI, the VLA and been established to provide a practical fore fixed space directions for a long future VLM-net on a routine basis. solution within the corning three years to period. This idea already has a long Absolute positions, proper motions trornstric history is therefore included in the programme stars: for a recent over- view see (A & A, 1992). Thus the HIPFARCOS mission will provide automatically a homogeneous and fairly dense stellar net of about 2.7 stars/sq. deg., maHy in the magnitude 0 Quasars BL- Lac 373 Intmal 7-1 0 (see Fig. 3). Furthermore, the Tycho Mission will 100 add some 500,000 falnter stars with pre- cise photometry although reduced as- trometric accuracy. However, If we re- call for example that already the AGK3 on the northern hemisphere and the CPC2 on the southern hemisphere pro- vide stellar densitlee of r 10 stars/ sq.deg. it Is obvious that the HIPPAR- 50 COS net shoutd swn be made denser and extended to much fainter limiting magnitudes by further catalogue pro- jects to keep up with the needs of large tetescopes with their small-fleld, highly sensitive area detectors. 2. Linking HIPPARCOS to the VLBl System Figure 1 : Magnit- distributbn of 4M radio mum for the Extragalactic Reference Fmme. The rigidly constructed HIPPARCOS stellar net still may contain a small un- known rotation of some madyear which will be reflected in the HIPPARCOS and parallaxes of those stars are ob- any previous fundamental catalogue, proper motions. After this residual mo- tained than in the primary extragalactic provided R can be linked in a unique way tion has been taken out, the HIPPAR- frame. At the same time these galactic to the VLBl based prZmary reference COS net will be adjusted to a fixed ori- objects can be easily accessed by opti- frame. gin, close to that of the FKUJ2000 sys- cal astrmetry because of their bright- Thus a multi-step approach is tem. Because none of the VLBl primary ness thus providing a natural link to the necessary for the construction of this sources is observed by HlPPARCOS di- optical spectral region. new reference frame, on the other hand, rectly, a number of indirect approach While being optimal candidates in the dealing again with stars which will re- has been worked out to provlde the Itnk radio domain, the primary swrces are flect kinematically their gatactic origin to th inerthl (rotation-ffee) extragatac- not vwy suitable to work with In the and the earth's and solar system's mo- tic system (see Table 1). With a view to optical spectral range directly. The tion, precise proper motions (p.m.) and the ongoing problems with HST, several reasons are obvious: firstly, the faint- parallaxes Rave to be determined. Be- ground-based, large observing pro- ness of their optical canterparts (most- cause of unavoidable systematic and grammes are undwway to provide thls ly beyond 18th magnitude, see Fig. I), random errors in this prom,the stellar link. (Fmchle and Kovalevsky, 1982; which requires large telescopes for ob- net will deteriorate substantiatly as a A & A, 1992; de Vagt et at., 1991). sewation, the astrometrically usable function of changing epoch. To maintain However, whereas the extragalactic field of which is << 1 degree, with an the initial high precision of the HIPPAR- VLBl net does not display a net rotation additional dramatic decrease of usable COS net and to improve the precision (2 because of the inertial nature of Its field size when CCDs are used instead mas/yr) of the HIPPARCOS proper mu- target objeEts, f he zero point ofthe R. A of the photographic plate. Therefore the tions, future astrometry satellite mis- system has to be adjusted, because object of interest and the reference aims are indispensable. Furthermore, a VLBl provides only R.A. differences, al- frame sources have to be very close in continuous improvement of the main as- though absolute declinatlorw. kd- the sky, e situation which will allow only tronomical constants as for example ing to the quoted IAU miutlon, the vsty occasionally to link any other object precession and nutation are of crucial R.A zero point shall be adjusted to the directly to the primary reference frame. importance to maintain a high-precision FK5/J2000 zero point at epoch J2000. Secondly, there is no redistic measur- reference frame. To achleve this goal, precise optical po- ing technique presently avallabb in the Although the selection of the abt sitions of a suitable subset of the radio optical domain to determine object po- 120,000 HIPPARCOS programme stars sources have to be determined in the sitions relative to these very faint prlma- was based primarily on astrophysical FK5 system. Furthermore, the physical ry sources over large arcs, contrary to proposals, the smning principle used nature of the radio and optical emission the radio. Therefore a practical reaIira- by the satellite and the requiremmt to and the morphology of the sources have tion of the new reference frame must be monitor continuously the satellite orien- to be studied in detail to ensure that the based in addition on a dense, global net tation by a large set of so-called star- optical and radio emission centres will of fairly bright stars. At the same time mapper stars forhrnately had the eonse- coincid to the precision of the VLBl this net should be of comparable preci- quence that the selection of the pro- measurements, Le.