The spectroscopic Be stars Atlas

A Ha survey of Be type stars

 Go to the Be stars atlas
Last update: November 30, 2008

Time series observations

2004 sessions

Gamma Cas (click here for enter)

Zeta Tau (click here for enter)

2001 sessions

Omega Ori          HD41511

HD206773 (1)      HD206773 (2)

 
3D view of dynamical spectrum of Omega Ori

Bottle spectral profile of some Be

About Be stars

The Be stars are defined to be rapidly rotating O, B, A-type stars of luminosity classes III-V which Balmer emission lines. Sometimes the line profiles of helium and iron also appear in emission. Many B stars show these characteristics, but the Be stars correspond only to the class of non-supergiants (classes III to V). Approximately 10% of the non-supergiant B-type stars present the characteristics of the Be stars. The majority of the Be stars belong to the spectral classes B0 to B7. 5% of Be are of spectral type O8-O9.5 and a small percentage (1%) belongs to the spectral type A0 or A1.

A characteristic of the Be stars comes from the variability of aspect of the hydrogen lines (and sometimes, helium and iron). Stars can on a cycle of a few years show lines in strong emission, usually completely absent (confused with the continuum) or in absorption as in a normal star. The frequency of the variations of  the profile lines covers a very broad range since some Be stars can show certain line-profile changes in a few hours or minutes whereas others can remain stable during years. Sometimes several cycles of variations with very different periods can be superimposed. The emission features themselves have a great variety of profiles. A typical situation is the shell configuration in which a narrow and deep absorption line is in the center of a broad emission line. Some time photospheric absorption line wings is visible on both sides of the emission line profile. It is probable that the majority of the Be stars present a "shell" phase to one moment of their life.

The origin of the emission of the lines is related to the presence of a disk and/or a flattened envelope in fast rotation. The circumstellar material rotational velocity, that can exceed 300 km/s, which explains why the majority of these stars show very widened lines (Doppler broadening). It should be remembered  that what is observed is an apparent speed, which differs from true the disk velocity V because of tilt of the axis of rotation i from the observer line of sight. Measured speed is thus V.sin i, the projected rotational velocity, and not V. If a Be star shows a low projected rotational velocity, it is most of the time because i has a  low value (polar-on view of the star). If the material in the disk is sufficiently optically thick, we observe a narrow absorption  lines from the part of the disk with is projected in front of the underlying star (for this situation, the material is moving essentially at right angle to the line of sight, so were is no Doppler broadening). This explains the Be-shell subclass.
 

The rapidly-rotation star with equatorial gazeous shell produces different line profiles (rigth) in function of the polar inclinaison (left). Up to down: 42 And (polar-on configuration), 4 Her, 1 Del, and 88 He (equatorial-on configuration). The last profile is a typical shell type: the star show emission wings and a sharp absorption core. Rotational model from A. Sletteback  (Space Sci. Rev., 23, 541, 1979). But note that for several Be stars the shell spectrum comes and goes frequently and it is difficult to understand why the Be phenomenon is so episodic on the basis of this simple rotational model. Other mechanism are presents...

It is generally assumed that it is the high spin velocity of the star which partly explains the presence of these circumstellar material, the force of gravity hardly compensating for the centrifugal force at the equator. However there is no absolute explanation to origin and the significant variations of the spectral characteristics of the Be stars. Concerning the abrupt changes in the spectral line profiles, on time scales of a few hours to a few days,  it is mentioned the presence of outbursts, like strong solar eruptions, and it seems that strong magnetic fields induced by dynamo action, large stellar wind, non-radial vibration or spots on the stellar surface plays a key role. Moreover, the Be stars often have a close companion, which must play a role because of phenomenon of mass transfer between the two components and can explain medium term variability. Clearly, Be stars are now classified as active hot star. But the exact mechanism explaining the short, medium and long term variations still remain to be found. The origin of the disk and the type of rotational velocity law of this disk is also not yet completely clear.

It is not a surprise to see that numerous Be stars are variable stars of small amplitude (0.01 to 0.1 mag.). In a recent article A.M Hubert and M. Floquet, Meudon Observatory (Investigation of the variability of bright Be stars using Hipparcos photometry, Astron. Astrophys. 335, 565-572 (1998) says:  The search for a link between spectroscopic and photometric variations, crucial for testing dynamic models of circumstellar matter, strongly suffers from lack of spectroscopic data available for last decades on Be stars. Carry out photometric and spectrometric observations is a very feasible spot for the astronomers amateurs with the tools available today!

At the present time, we don't know how a B star can become a Be or a Be-shell star and we don't know if all the B stars will become, one day, Be stars. All this explains the interest to study these peculiar stars. This is a very dynamic subject of research in present time and observations are urged for constraining and consolidating numerical models of this mysterious objects.

Various form of the H-alpha spectral profile for Be stars
(1 Del, 10 Cas, Omicron Cas, 42 And, HD21620, 88 Her, Omicron And, HD21455)

About the Atlas

The study of the Be stars is a long tradition for French amateurs, since a program was undertaken in 1992 with the  60-centimeter telescope at Pic du Midi (T60) by Alain Klotz, Valerie Desnoux, Daniel Bardin and some other volunteers. This pioneer work clearly showed an example of a research program accessible to the amateurs and useful for the professionals.

My own observations of the Be stars correspond to a long-term monitoring of the H-alpha emission-line (and some time of the He I line at 6678 A), extending over several years. The aim is to collect a large amount of homogeneous data of this spectral region. The main telescopes used are a Celestron 11 (0.28m aperture F/10, noted T280 in the data base) and a Takahashi CN-212 (0.21m aperture F/D=3.5, referenced T212 in the atlas data). I have also used 190-millimeter Lichtenknecker flat-field camera at F/D=4 (referenced T190), a Takahashi FS-128 refractor (128-mm aperture, referenced R128), an AstroPhysics StarFire 120 refractor (120-mm aperture, referenced R120) and a Takahashi FSQ-106 refractor (106-millimeters aperture, referenced R106). The CCD camera are Audine models and the CCD are Kodak KAF-1602E and KAF-0402ME. The pixel size is of  9 microns. The actual more resolving spectrograph used is LHIRES2 (R=17 000, spectral sampling of 0.115 A/pixel). I used also spectrograph having a resolution power of about R= 7 000 (one pixel corresponds to 0.3805 A/pixel at H-alpha region) and R=10 000 (0.2293 A/pixel - since June 2003).

Observations are realized at Castanet-Tolosan (near Toulouse, south of France, Longitude: 1°30'30" Est, Latitude 43°31'03", Altitude: 157 m) and some time, atop the Pic du Midi Observatory (at the 0.6m telescope).

Images and spectra are processed with Iris and VisualSpec software. The main steps of the reduction scheme were: correction of the raw data for dark signal and bias signal, flat-field correction, sky subtraction, wavelength calibration (use of H2O vapor telluric lines near H-alpha, and some time O2 at 6872 A) and normalisation to continuum intensity. In order to increase the signal-to-noise ratio, several bias, darks and flats are recorded and added before use (10 typically). The spectral calibration precision is ±0.2 angstroms typically at R=7000 (±0.05 A with the R=17 000 spectrograph - correlation method by the use of H2O vapor telluric lines). For the moment the wavelength in the atlas are not corrected for the radial velocity of the Earth. Note also that the time is the UT (geocentric, not heliocentric for the present reduction). Continuum normalization was performed in two steps: first the instrumental spectral response was removed (the instrumental spectral efficiency is determined by comparison of the measured profile for A0V stars - like Vega - with the know standard profiles (Pickles' atlas)). Second, remaining gradients were removed by fitting the continuum by a cubic polynomial function. Finally, the telluric H2O lines are removed (the telluric lines model is from the GEISA data base). This last procedure is applied regularly since August 2000 (all the data base was re-processed later). The detailed processing pipeline is described here. Characteristics Be parameters extraction is described here (equivalent width, V/R, ...).

I have merged in my database unpublished observations made at the T60 between 1992 and 1997 (with authorization) in order to be used for a very useful long term survey evolution.

Some spectra are from Stephane and Didier Morata who observe with a 300-mm telescope near Martigue (France) (see: http://perso.wanadoo.fr/sdmorata/). The resolution of these spectra is slightly weaker than my actual spectrograph. The resolution power is estimateded to 2.5 A and the reciprocal spectra sampling is 1.05 A/pixel. Morata's spectra are identified by the SDM indicateur.

Two examples of Be stars spectra around H-Alpha (R=3000 spectrograph). Up: 31 Peg. Down: HD193182.