Transcript aas_gdemessieres - Astronomy at Swarthmore College

Chandra Emission Line Diagnostics of t Sco
Geneviève de Messières (Swarthmore College ‘04), Carolin Cardamone (Wellesley College ‘02), David H. Cohen (Swarthmore College),
Joseph MacFarlane (Prism Computational Sciences), Stanley Owocki & Asif ud-Doula (Bartol Research Institute, University of Delaware)
The quality of astrophysical X-ray spectra has improved dramatically in the last decade: The high-resolution of
the Chandra HETGS allows us to study the profiles of spectral lines in t Sco to determine the velocity structure
of the X-ray-emitting material, and line-intensity ratios to determine its height above the photosphere.
The processes by which hot stars produce Xrays are not yet fully understood
- Cooler stars like the Sun generate X-rays through
magnetic confinement and heating of a corona. However,
hot stars are generally thought to lack the convective
envelopes and magnetic fields assumed to be necessary for
coronal X-ray production.
- The leading theory for hot stars X-ray production is
shock-heating in radiatively-driven winds. The line-force
instability (Owocki, Castor, Rybicki 1988; Feldmeier 1995) is
a natural mechanism for generating shocks in a radiationdriven wind.
- A third, hybrid, mechanism is the magnetically
confined wind shock (MCWS) model (Babel and Montmerle
1997). A star with a magnetic field and a substantial
radiation-driven stellar wind can develop a disk of colliding
plasma around its magnetic equator that would generate
shock-heated plasma and associated X-rays. Asif ud-Doula is
discussing dynamical modeling of the MCWS mechanism at
these proceedings (Poster 135.07).
t Scorpii
B0 V, mv= 2.8
D = 132 pc
Teff = 31,400 K
Vsini = 20 km s-1
ROSAT (1993) spectrum of t Sco
M ≈ 10-9 to few X 10-8 Msolar yr-1
v∞ = 1500 km s-1
Age ≈ 1 Myr
Chandra (2000) MEG spectrum (85 ksec)
ASCA (1997) spectrum
Line Widths Provide Velocity Information
There is only modest broadening of t Sco’s
spectral lines. A fit of a delta-function lineprofile model of the Ne X line shows that the
actual width exceeds the instrumental
broadening, but only by several 100 km s-1.
Redshifted as well as blueshifted UV
wind absorption from Copernicus
Line Ratios Provide Information About Location
- The physical picture that emerges from these diagnostics suggests very hot
plasma that is a moderate distance from the star (1 to 2 R* above the photosphere) and
that is not moving very fast. Some of it, in fact, appears to be falling toward the star.
We can resolve three closely spaced lines in each helium-like
triplet: the resonance (R), intercombination (I), and forbidden
(F) lines. The ratio of the F to the I line is a diagnostic of the
strength of the ultraviolet field, and therefore distance from
the star (because the strength of the UV field is directly related
to radius through the dilution factor).
O VII
Ne IX
Discussion
- This is inconsistent with both the coronal and line-force instability models.
(The material is too far from the star to be explained by a coronal model, but is moving
much more slowly than a stellar wind driven by line-force instability.)
In a strong UV field,
electrons can be
excited out of the
metastable upper
level of the forbidden
line (3S  3P) before
they spontaneously
de-excite, weakening
the F line in favor of
the I line.
- A possible alternative: The magnetically confined wind shock model model
naturally explains how wind plasma could be shock-heated at a significant distance
from the star, while remaining relatively stationary.
Density
Y- Velocity
WINDRT simulations of neon IX
(Left) The Ne X Ly-a line from the spectrum of t
Sco in comparison to lines from Capella (which
is a cool coronal source) and z Puppis (which has
a strong stellar wind with embedded shocks).
laboratory rest-wavelength
Roban Kramer has studied the broadened line profiles of
stars with stronger winds (Poster #135.13).
-1000
Observed range for tau Sco
Mg XI
WINDRT simulations (MacFarlane et al 1993; MacFarlane, Cohen, & Wang 1994)
model the UV field strength necessary to cause the observed F/I ratios.
Si XIII
Our results: Lines in t Sco are not as narrow as those in coronal sources, but
they are only slightly broadened. Typical velocities are 200 km s-1, while the
observed wind terminal velocity for t Sco is at least 1500 km s-1. Furthermore,
the lines are slightly redshifted (mean value ~50 km s-1 for ~20 lines).
http://hven.swarthmore.edu/~cohen/group/AAS2002_GdM/
Ion
oxy gen VII
neon IX
magnesium XI
silicon XIII
F/I
0
0.0845
0.3057
2.1623
Error
0.01 (est.)
0.0214
0.0654
0.286
Range of radii (r/R)
<5
2.2 - 3
1.8 - 2.5
1.1 - 1.5
Our results: We found that the hot plasma is located between
1.5 and 5 stellar radii from t Sco, with the majority located at
about 2-3 radii. The Si XIII data suggest that the hottest
material is nearest the star, but there are uncertainties in the
photospheric EUV flux that may affect this result.
vy (km/s)
1000
The MCWS model: A large-scale magnetic field can channel ionized wind material
toward the magnetic equator, where it collides with material from the opposite
hemisphere, leading to a strong standing shock and generating X-rays. (See udDoula & Owocki, poster 135.07.)
t Sco is an unusually young star (Kilian 1994), and it could retain a primordial magnetic field from its
formation, thus it is a potential candidate for the MCWS mechanism. A polar magnetic field strength
of just a few 10’s of G in this star could lead to significant magnetic effects (h* ~ 1).
Preliminary analysis indicates a significant quantity of plasma with temperature in excess of 107 K,
hotter than the X-ray emitting plasma seen in most OB stars, and potentially explained in the context
of strong magentic wind shocks (much like q1 Ori C).
Line centroid redshifts might be explained if the shock-heated plasma arises from the interaction of
density condensations in the wind, falling back toward the star, as has been modeled in this star by
Howk et al. (2000).
The periodic infall of wind condensations is an inherent property of the MCWS model
(‘magnetospheric emptying’). See simulations by ud-Doula and Owocki (Poster 135.07).
Presented at the American Astronomical Society winter meeting - Washington, DC - January 6-10, 2002