Astron. Astrophys. 349, 169-176 (1999)

1. Introduction

Mass loss is of central importance in stellar evolution. For intermediate and low mass stars mass-loss becomes significant when they enter the red giant phase (Schaller et al. 1992). Theoretically, mass loss from M giants is not yet understood. An important factor that hampers the development of theoretical models is the paucity of observational constraints as to what is driving the winds of M giants. The empirical determination of their wind structure is therefore fundamental.

Vogel (1991), Vogel et al. (1992), and Pereira et al. (1999) analyzed the M-giant wind in the eclipsing symbiotic systems EG And and SY Mus. In these systems, a hot radiation source probes the column density of neutral hydrogen of the wind from the red giant due to the Rayleigh scattering effect. Assuming spherical symmetry for the mass loss, , they derived empirically the velocity structure of the wind of the M giant.

In this paper we draw attention to two complicating factors which have been neglected so far in this type of analysis. Based on a new orbital solution of SY Mus (Sect. 3) we show that the UV eclipse curve is not symmetric with respect to mid-eclipse. We investigate the density structure and derive the mass-loss rate of the red giant (Sect. 4). We then use HST spectra of RW Hya and demonstrate, that in addition to Rayleigh scattering, an important source of absorption is due to line blanketing (Sect. 5).

© European Southern Observatory (ESO) 1999

Online publication: August 25, 1999
helpdesk.link@springer.de