Transcript Variability of Be stars

Variability of Be Stars:
A Key to the Structure of their
Circumstellar Environments
Anatoly Miroshnichenko
University of Toledo
Variable Star Meeting 2004. BGSU, April 3
In Collaboration with
Karen Bjorkman
Jon Bjorkman
Alex Carciofi
University of Toledo
Outline
•
•
•
•
•
Brief history of studies of Be stars
Basic properties of Be stars
Variations of the spectrum and brightness
How can we use the variability to get the
physics?
Current state of the research
Classical Be Stars
First discovered group of emission-line stars
Emission lines in the spectrum of  Cassiopeae
were found in 1867 by visual spectroscopy
~200 Be stars are currently known among 1660
B-type stars brighter than V=6.5 mag
Main properties of classical Be stars
• Non-luminous rapidly-rotating objects displaying
emission-line spectra
• Emission line profile shapes are usually double- or
single-peaked at a low or moderate spectral resolution
• Infrared (IR) radiation excesses
• Polarization of the continuum radiation
• Active emission-line phases may last for decades and are
followed by no-emission or shell-line phases
• Metallic line profiles (e.g., Fe II) suggest that the
circumstellar gas is involved in a Keplerian motion
around the star with small radial velocities (a few km s-1)
Basic Stellar Parameters
Some H profiles
IR Excess
1998
 Aqr
1983
Polarization
2.0
1.8
1.6
1.4
P (%)
1.2
1.0
0.8
0.6
0.4
 Tau
0.2
0.0
2000
4000
6000
 (Å)
8000
10000
Circumstellar Disks
Origin of the Observed Features
Line emission – ionized circumstellar gas
IR excess – free-free emission
Polarization – Thomson scattering
The polarization spectrum and spectral line
profiles imply a flattened, disk-like, envelope
Theories of the Be phenomenon
Elliptical disk model (Struve 1931, ApJ, 73, 94) Keplerian rotation of particles in a circumstellar disk.
No explanation for the disk long-term stability.
Rotation-pulsation model - changing inflow and outflow
superposed onto the rotational motion in the disk. Variable
stellar wind as triggering mechanism for the V/R
variations (Doazan et al. 1987, A&A, 182, L25).
No explanation for the disk formation.
Theories of the Be phenomenon
Wind-compressed disk model (Bjorkman & Cassinelli
1993, ApJ, 409, 429) - a rotating wind produces a disk.
Assumption: the outward acceleration is smaller than the
rotation  the material will orbit down to the equator
before it is accelerated outwards.
But: small non-radial forces act against disk formation.
A disk can be produced by mass transfer in binary systems
(Kriz & Harmanec 1975, BAICz, 26, 65), where the mass
gainer spins-up to critical rotation.
Can explain formation of 20-40 % of all Be stars (Pols et
al. 1991, A&A, 241, 419) or even less (~5%, Van Bever &
Vanbeveren (1997, A&A, 322, 116).
Theories of the Be phenomenon
Non-Radial Pulsations may be a triggering mechanism
for the mass loss from at least early-type Be stars
(Rivinius et al. 2003, A&A, 411, 229)
Modeling the Be Star Disks
Input parameters:
1. Stellar (Teff , log g , L)
2. Circumstellar: 0 ,Te(r) , geometry, density
distribution (=0 r-n)
Observational data:
1. Spectral energy distribution
2. Spectral line profiles
3. Polarization spectrum
Typical Modeling Results
Disk opening angle - a few degrees
Statistical studies suggest opening angles from 5
to 25-40 degrees
Density at the disk base - ~10-11- 10-12 g cm-3
Density distribution slope - 2.5-3.5
Theoretical Disk Structure
from Carciofi & Bjorkman (2004, Polariz. Conf.)
Modeling Results for  Tauri
from Carciofi & Bjorkman (2004, Polariz. Conf.)
Problems with Snapshot Modeling
Theoretical (simplified assumptions):
Smooth density distribution
Uncertain disk size
Parameter space degeneracy (0 – disk scale height)
Observational:
Non-contemporaneous data
Limited spectral range
Line Strength Variations
Complex Profiles
 Oph
Variations
 Aquarii: H Variations
What can We Get from Variability?
Reveal the true basic stellar parameters and
content of the system
Determine the circumstellar contribution to the
continuum
Learn about the mass loss history and mass
distribution in the disk
Be star spectroscopy at the
Ritter Observatory
• 1-meter telescope with a fiberfed echelle spectrograph
and a 1150x1150-pixel CCD in the Coude focus
• 9 non-overlapping orders, 70 Å each, range 5285-6600 Å.
Includes spectral lines of FeII 5317 & 6383, HeI 5876,
NaI 5889 & 5895, SiII 6347 & 6371, and H
• Spectral resolving power R (/) ~ 26000
• Spectra of stars brighter than 7.5 mag can be obtained in 1
hour with a signal-to-noise ratio of ~100
• ~1700 spectra of ~ 45 Be stars obtained in 1991-2004
Ritter data statistics (as of 2004/03/26)
Name
 Cas
 Lyr
93-95 96 97
3
3 23
81
98
14
36
99
28
44
00
40
17
01
50
02
38
03
22
04
5
Tot
221
177
 Dra
1
8 16
29
19
5
34
42
26
1
180
 Tau
 Aqr
6
2
20
8
31
30
15
14
19
23
10
7
25
19
4
143
126
 Sco
 And
 CrB
1
30
1
20
13
1
40
1
8
45
41
1
10
2
149
70
66
9 11
5
8
8 23 22 20
3
36 201 176 210 161 324 273 223 16
58
76
1709
13
8 17
 Per 5
 CMi
Total 89
3
10
8
14
26
1
11
2
8
8
Some results from Ritter program
• Discovery of 2 new Be stars, HD 4881 (V=6.2, B9 V) and HD 5839
(V=6.7, B9 V) (Miroshnichenko et al. 1999, MNRAS, 302, 612).
• Periodic RV variations (84.3 d) of the emission peak and absorption
wings of the H profile in  Aquarii are discovered during the
normal B-star phase (Bjorkman et al. 2002, ApJ, 573, 812).
• Periodic RV variations (205 d) in the H line of  Cassiopeae
(Miroshnichenko, Bjorkman, Krugov 2002, PASP, 114, 1226).
• New orbital solution for the  Scorpii binary and monitoring of the
initial disk formation stages (Miroshnichenko et al. 2001, A&A,
377, 485; 2003, A&A, 408, 305).
 Cassiopeae
 Aquarii
 Aquarii
 Scorpii
 Scorpii
 Scorpii
Conclusions
• Long-term multi-wavelength and multi-technique
observational studies are needed
• Such observations give us information about:
- the disk structure
- the system content and fundamental parameters
- the mass loss origin and evolutionary state
• Most of the needed observations can be obtained
with relatively small telescopes