Astron. Astrophys. 350, 485-490 (1999)

1. Introduction

In a previous paper we reported on the discovery of an H burst in the B emission line star HD 76534 (Oudmaijer & Drew, 1997). During an observing run in 1995, two days after a strong H emission line was observed, the star was re-observed, and found to have only a photospheric absorption line. Spectacularly, two hours later, the line was again in emission, still increasing its strength with respect to the continuum.

Especially marked spectral variations have been detected in a few other Be stars, notably µ Cen (Peters, 1986; Hanuschik et al. 1993, Rivinius et al. 1998) and Eri (Smith et al. 1991, Smith et al. 1997a+b). These two stars have long periods with relatively stable H emission, and sometimes undergo an H outburst where the emission grows to a maximum within days, displaying rapid variations of the violet and red peaks of the H line. The emission then fades on longer timescales.

In contrast to both µ Cen and Eri, the observed timescale of the H outburst of HD 76534 was an order of magnitude shorter, while the line profile did not show any V/R variability. Instead, the line that was present only two hours after the absorption was observed, was similar in profile and V/R ratio to existing high resolution H spectra of the object, and did not betray any signs of on-going formation of recently ejected disk material.

Two hypotheses were put forward by us to explain the rapid variations. In analogy with the hypothesis for µ Cen, a sudden burst of mass loss was first considered (see Hanuschik et al. 1993, Rivinius et al. 1998), but based on the above arguments it was discarded in favour of the idea that a stable rotating Keplerian disk was already present around the star, but that a lack of ionizing photons from, for example, the stellar photosphere failed to produce sufficient ionizations and subsequent recombinations to push the H line into emission. From simple considerations, it was found that a slight change in ionizing flux can indeed ionize an existing stable neutral disk, and result in detectable H emission.

It is not clear however where the change in ionizing radiation should come from. By analogy with the EUV variations of the Cep star CMa (Cassinelli et al. 1996), it is possible that stellar pulsations are responsible for this behaviour. CMa shows relatively large (30%) variations in its Lyman continuum, which are not as readily visible in the optical (Cassinelli et al. 1996). A similar effect could be happening in the case of HD 76534; at the times when the Lyman continuum is at minimum, no H emission is visible, while at maximum the line will develop. A critical test of the stellar pulsation hypothesis would be to monitor the star for several hours up to several days to investigate whether any periodicity would be present in the H emission of the object.

On the other hand, Smith et al. (1997b) reaches a similar conclusion to explain, amongst other phenomena, the H variations in Eri. A source of extra Lyman continuum photons could be responsible for extra ionizations and recombinations in the circumstellar material. Smith et al. (1997b) find that this can be explained by the occurrence of heated slabs, possibly related to magnetic activity, close to the stellar photosphere. This activity does not appear regularly, so no apparent periodicity in the H line, certainly not on timescales of hours, would be expected.

Previous to Oudmaijer & Drew (1997), observations of the H line of HD 76534 were reported only twice, by Thé et al. (1985) and Praderie et al. (1991). Their spectra were taken one year apart, and showed `indistinguishable' (cf. Praderie et al. 1991) line profiles. We measure an equivalent width (EW) from Thé et al. of -7 . The exposure times used by these authors were 2 and 2.5 hours respectively, so that any shorter term variations in either the line profile or EW would have been washed out.

Since our previous paper, several new datasets on HD 76534 have been published. Recently, Corcoran & Ray (1998), report a measurement of the EW of the H of -14.3 , on a spectrum that was obtained in the last week of 1991 (their only Southern observing run) and Reipurth et al. (1996) show a spectrum, obtained in February 1993, with an EW of -10 . These two papers do not mention the exposure times. Oudmaijer & Drew (1999) report on observations taken in December 1995 and December 1996 and found an EW of -6 and -4 respectively (we note a typographical error in their Table 2, the data for 30 December 1996 and December 1995 should be interchanged). It is clear that the H line of HD 76534 is variable on long timescales.

The main aim of this paper is to present time-resolved observations of the object at high resolution as a first step in determining whether H is prone to any variations on timescales of minutes. The observations thus serve as a check whether the long integration times by Praderie et al. (1991) and Thé et al. (1985) would indeed have washed out any short-term variations.

This Research Note is organized as follows: In Sect. 2 we present the time-resolved observations of the object obtained with the UCLES spectrograph on the AAT. In Sect. 3 we will search for variations in the data and revisit the multi-epoch V band photometry obtained by Hipparcos. We will conclude in Sect. 4.

© European Southern Observatory (ESO) 1999

Online publication: October 4, 1999
helpdesk.link@springer.de