Be stars in open clusters - II. Balmer line spectroscopy
Astron. Astrophys. Suppl. Ser.
Be stars in open clusters
J.M. Torrejon 1 2
J. Fabregat 1 2
G. Bernabeu 0 1
S. Alba 1 2
0 Departamento de Ingenier a de Sistemas y Comunicaciones, Universidad de Alicante, Apdo. 99, 03080 Alicante , Spain
1 Send o print requests to: J.M. Torrejon
2 Departamento de Astronom a y Astrof sica, Universidad de Valencia , 46100 Burjassot, Valencia , Spain
We present H line pro les (R ' 3 103, the photospheric absorption lines are lled-in with an unS=N ' 102 at continuum level, ' 50 for some objects) known amount of circumstellar emission in such a way for 58 stars in 8 northern open clusters and associations, that it is very di cult to deduce either the spectral type complemented with H and Hγ pro les for 36 of them as of the underlying star or the actual amount of the emitwell as H and H for 16 programme stars. Our observa- ting material, the line pro les themselves are distorted and tions, though lower in resolution, yield results which are the photometric indices contaminated. As a consequence in good agreement with those obtained in earlier works we cannot apply the usual photometric and spectroscopic based mainly on eld Be stars. Our data suggest a similar techniques to derive the astrophysical parameters of the spectroscopic behavior for both eld and cluster Be stars. underlying star.
lines; pro les | stars; emission line; be | open clusters and associations; general
1. Introduction
Be objects are de ned as stars of spectral type B, and
luminosity classes III to V, whose spectra show, or have
shown at some time, Balmer (and sometimes other) lines
in emission. The origin of this emission is atomic
transitions in a circumstellar envelope, relatively cool ( 104 K)
and relatively dense ( 1011 cm−3) with a composition
similar to that of the stellar photosphere. The contribution
of the circumstellar envelope has dramatic e ects upon
the observables: the slope of the continuum is distorted,
Our goal is to elaborate a method to determine the
intrinsic parameters of Be stars from uvby photometry
and Balmer line spectroscopy. A rst attempt to develop
such a method was the work of
Fabregat & Reglero (1990)
based on published studies of emission properties of
circumstellar disks around Be stars. However the
photometry and the spectroscopy were not simultaneous. Since Be
stars are known to be highly variable some uncertainty
is expected in the derived relationships. Thus, the latter
authors consider their calibration to be preliminary.
With this in mind we planned to build a set of stars
with high quality photometry and spectroscopy, as
simultaneous as possible and with accurate values of
reddening and distance modulus. This latter requirement
was ful lled by observing Be stars with well established
membership of open clusters so that both interstellar
reddening and distance modulus can be inferred from
that of the cluster
(Mermilliod 1982; Slettebak 1985)
. In
Fabregat et al. (1996)
, hereafter called Paper I, we
presented the photometric measurements along with a
careful analysis of the data. One of the main results obtained
was the observation of a clearly di erent photometric
behavior between Be stars earlier than B5V and later ones.
This fact, of course, is expected to contribute to the
uncertainty of the above mentioned calibration.
In the present paper we present the spectroscopy and
analyze the data as far as the available resolution makes it
possible, mainly in order to guarantee the homogeneity of
the sample. Further work will include the elaboration of
a calibration valid for Be type stars, the determination of
astrophysical parameters for circumstellar envelopes and
their modeling.
2. The observations and reduction procedure
We carried out two observing runs during the months
of December 1992 and August 1993, both with the
Cassegrain spectrograph attached to the 2.2 m telescope
at the Calar Alto Observatory, Almer a, Spain. The
detector used in both runs was the GEC#15 CCD. The two
instrumental arrangements were di erent, giving slightly
di erent resolutions. For the rst con guration, we used
the f =3 camera, at 0.66 A/pixel spectral dispersion,
giving a resolution of about 1.5 A at 6000 A. For the latter
arrangement we used the f =1:5 camera which produces a
spectral dispersion 1.36 A pixel−1 giving a resolution of
about 2.7 A, at 6000 A. The red spectra were centered
on the H line at 6563 A. The blue spectra included only
H (4861 A) and Hγ (4340 A) in the rst run, while in the
second one we extended the range from H to the Balmer
discontinuity ( 3647 A).
The spectra were reduced by means of the Starlink
FIGARO software package (Shortridge 1991). Several bias
frames were taken during the night, their stability was
checked and then they were averaged, before subtraction
from the program frames. No drift in dark current was
found after several hours of operation, and at di erent
time exposures. The flat elds were recorded at the end
of the night. A smooth functio (...truncated)