1 Introduction

The Hyades is the nearest rich open cluster and as such has played a fundamental role in astronomy as a first step on the cosmological distance ladder and as a test case for theoretical models of stellar interiors (Lebreton 2000). From the first use of the converging point method by Boss (1908) up to the use of pre-Hipparcos trigonometric parallaxes by van Altena et al. (1997), an important goal in astrometry has been the determination of an accurate distance to the cluster. With the advent of the Hipparcos Catalogue (ESA 1997) the Hyades lost its unique status for distance calibration, but as the depth and internal velocity field of the cluster were well resolved by Hipparcos, focus could instead be turned to its three-dimensional structure and kinematics (Perryman et al. 1998). A deeper understanding of the dynamics and evolution of the cluster should now be possible through detailed comparison with N-body simulations.

Thanks to the accurate Hipparcos measurements, the Hyades has recently acquired a completely new role as a practical standard in observational astrophysics: it is one of very few objects outside the solar system for which the accurate radial motion can be determined by geometric means, i.e. without using the spectroscopic Doppler effect. From a combination of Hipparcos parallaxes and proper motions, Madsen et al. (2002) obtained "astrometric radial velocities" for individual Hyades stars with a then estimated standard error of about 0.6 km s^-1. Currently the Hyades is the only cluster for which astrometric radial velocities are derived with individual accuracies better than 1 km s^-1, but the technique may be extended to many more objects with future space astrometry missions (Dravins et al. 1999b).

Astrometric radial velocities are important mainly because they make it possible to determine the absolute lineshifts intrinsic to the stars, through comparison with spectroscopic measurements. Such lineshifts are caused for instance by convective motions and gravitational redshift in the stellar atmospheres (Dravins et al. 1999a). Absolute lineshifts could previously only be observed in the solar spectrum, but are now within reach for a range of spectral types through the use of astrometric radial velocities. The present paper is part of a research programme at Lund Observatory in which absolute lineshifts are determined and used as a diagnostic tool in stellar astrophysics (Dravins et al. 1997, 1999b; Lindegren et al. 2000; Madsen et al. 2002; Gullberg & Lindegren 2002).

A major uncertainty in the astrometric radial velocities originates in the internal velocity dispersion of the cluster, which limits both the accuracy of the cluster motion as a whole, and that of the individual stars. A primary goal of the present investigation is to find out whether a better understanding of the internal velocity structure of the cluster, obtained through N-body calculations, can be used to improve the accuracy of the astrometric radial velocities.

Section 2 briefly recalls the kinematic information, including astrometric radial velocities, that can be derived from Hipparcos data. Section 3 describes the model used to simulate the evolution of the cluster up to its present state, and its subsequent observation, as well as the main properties derived from the simulations. Implications for the accuracy of the astrometric radial velocities are discussed in Sect. 4, followed by a discussion of non-modelled effects in Sect. 5, and conclusions.