Transcript UVObs - DU Portfolio

Epsilon Aurigae
Ultraviolet observations
Prepared for Eps Aur team meeting
April 29, 2013
Kathy Geise
Research Objectives
• Differential analysis of Eps Aur system
• Characterize the F star out of eclipse
• Identify spectral features
• Adjust to stellar frame - completed
• Third party application for line identification
• Line list from NIST
• EW, RV
• H alpha – emission and absorption fitting
• Characterize polarization features
• Degree and position angle
• Line polarization vs. broadband intrinsic polarization
• Remove interstellar polarization
Questions
• What is the nature of the broadband polarization in eps Aur
system out of eclipse?
• Broadband (continuum) polarization is variable
• Electron scattering?
• What are the relevant position angles?
• Is broadband polarization behavior related to Hα line
polarization behavior?
• Position angle Hα absorption not the same as metallic lines
• What is the source of Hα emission in eps Aur system?
• Does eps Aur F0 star have an active (i.e. variable)
chromosphere?
• Is chromosphere asymmetric?
• Source of Hα emission?
• (Ferluga & Hack (1985) consider Hα emission chromospheric)
• Source of broadband polarization?
Ultraviolet observations
• Why observe stars in the ultraviolet?
1.
2.
Detect emission lines formed in the outer atmosphere layers
(chromosphere, corona, wind, and disk) against the weak
photospheric background for stars later than early-A
Most resonance lines are at UV wavelengths
• Important examples: H I, C I-IV, N I-V, O I, Mg I-II, Si I-II, S I-II, Fe II
3.
4.
5.
6.
Hot companions dominate the UV emission of cool stars
Wind signatures for hot stars are best studied using UV lines
Important density sensitive line ratios are in the UV (e.g., O IV
1400A multiplet)
Spectral lines of ions that are formed at temperatures hotter
than about 8,000 K are located in the UV (e.g., C IV, Si IV, N V)
http://jila.colorado.edu/~jlinsky/lecture02.html
Ultraviolet observations
• Why do we need high spectral resolution?
1.
2.
Measure a weak continuum against a noisy background
Confusion due to overlapping lines
• Especially for metal rich stars with molecules
• Fe II lines are everywhere
3.
4.
Accurate line widths, shapes, identify and separate multiple
components
Determine whether lines are optically thick
• Essential for abundance analyses (need to resolve the line)
• Existence and column density of interstellar and circumstellar lines
(often need 3 km/s resolution or better)
UV spectra – IUE
• International Ultraviolet
Explorer (IUE)
•
•
•
•
•
Ultraviolet spectrophotometry
High resolution (0.1 - 0.3 Å)
Low resolution (6 -7 Å)
Range 1150 Å to 3200 Å
January 1978 to September
1996
• Scattered light at short
wavelengths
• 681 observations of Epsilon
Aurigae from 1978 – 1996
• (196 high dispersion, 330 low
dispersion large aperture)
• http://archive.stsci.edu/iue/
UV spectra – UVS
• Ultraviolet spectrometer
(UVS) aboard Voyager I
spacecraft
• Ref. Altner et al. (1986)
• From (R. Polidan, private
comm.)
• See papers Holberg, J.
• Range 535 to 1702 Å
• One observation February
1984
• Spectrum consistent with
B5 star
• Possibly misidentified?
• http://vega.lpl.arizona.edu/voyager_uv
s/instrument.html
UV spectra - FUSE
• Far Ultraviolet
Spectroscopic Explorer
(FUSE)
• Launched June 24, 1999
• Decommissioned on
October 18, 2007
• Two silicon carbide (SiC)
gratings optimized for 9051100 Å bandpass
• Two lithium fluoride (LiF)
gratings optimized for 10251187 Å
• R ≈ 10,000 – 20,000
• One observation of Epsilon
Aurigae on 2001-01-07
UV spectra - GHRS
• Goddard High
Resolution
Spectrograph (GHRS)
aboard HST
• Resolution 0.6, 0.06,
0.012 Å
• 1050 to 3200 Å
• Decommissioned 1997
• One observation of
Epsilon Aurigae on
1996-02-16
UV spectra - COS
• Cosmic Origins
Spectrograph (COS) on
HST
• Far-ultraviolet (FUV)
detector 900 to 2150 Å
• Near-ultraviolet (NUV)
detector 1650 to 3200 Å
• Resolution (R ~ 20,000)
and (R ~ 3,000)
• Three observations of
Epsilon Aurigae
• 2010-09-01
• 2010-12-09
• 2011-03-18
Broad absorption feature
2795.528 Å
0.0 eV – 4.43 eV
2S ½ - 2P0 3/2
2802.704 Å
0.0 eV – 4.43 eV
2S ½ - 2P0 ½
Wavelength Å
Fig. 3. Mg II doublet emission (c. 10,000 K) nearly unaffected by eclipse;
P Cygni profile (wind) or interstellar absorption; Mg II emission not
normally seen in F-type supergiants; spectra from IUE Chapman et al.
(1983) also Chapman (1985) for comments.
Eaton et al. (2008)
Eps Aur
Orbit Neptune ≈ 30 A.U. ≈ 6500 Rsun
1 RSun = 0.0046 A.U.
Harper et al. (2005)
ζ Aurigae eclipsing binary system
TBr
15000 K
300 K
5K
Fig. 5 Three dimensional hydrodynamic simulation. Harper et al. (2005)
Fig. 8. Model of the ζ Aurigae system from Chapman, R. (1981)
Fig 5. Chapman
(1981)
Mg II resonance
lines ζ Aurigae
A.Circumstellar
absorption
B.Interstellar
absorption
C.Emission
Εpsion Aurigae C II (near 1335 Å) emission profiles from
HST/COS 2 Sept 2010 observation. Slow wind (100 km s-1)
Howell et al. (2011)
http://adsabs.harvard.edu/abs/2011AAS...21725707H
IUE low resolution
Boehm et al. (1984)
Out-of-eclipse far UV continuum variability and UV excess
Same epochs, longer wavelengths
Fig 3.
Monochromatic UV
light curves.
The O I (1304 A)
curve seems to be
unaffected both by
the eclipse and by
intrinsic activity
Boehm et al. (1984)
Fig. 4. Corrections for scattered light. Kurucz (1979) model Teff = 7500 K, log
g = 1.0, log A = 0.0 (thin line). Filled squares are scaled data for Canopus.
The excess UV may not be attributed to scattered light or F-star. Altner et
al. (1986)
UV variability
Table 3. Amplitudes of Brightenings during totality
Wavelength [A]
1st Brightening *
(JD 2445400)
3000
2nd Brightening **
(JD 2445700)
0.23 mag.
2900
0.10 mag.
0.26
2800
0.13
0.26
2700
0.13
0.30
2600
0.16
0.30
*No confirmed brightening in the optical during this epoch
** Brightening 0.03m (U), 0.18m (B), 0.11m (V)
Kang, Y. –W. (1990) from IUE data
Fig. 1
21 January 1983
0.4m brightening
in B band of
duration about
20 minutes was
attributed to
flare.
Nha, I. & Lee, S.
(1983)
Fig. 4. Deepening of the Mg II 2800 A resonance doublet wings during eclipse.
Out-of-eclipse (solid line) and scaled eclipse data (dotted line). Features align near
2780 A and 2820 A. Additional broad absorption during eclipse or an effect of
“filling” by out-of-eclipse activity. Ferluga & Hack (1985)
Ly α - geocoronal
Red emission peak
Blue emission bump
Middle absorption feature
GHRS spectrum
•
•
•
•
•
Observed before secondary eclipse (1996 February 16)
R = λ/Δλ ≈ 2,000
Corrected for reddening E(B – V) = 0.3
Sloping continuum higher in the red
Identified emission lines are transitions to ground and low excitation
states of abundant atoms and ions
• Unresolved in IUE are four (of five) lines between 1302 and 1309 Å
(see Table 1)
• Originate near uneclipsed secondary or in chromosphere-like region
around the primary
• Lines are emission only, emission with absorption components and
absorption only
• Mg II (1239.93 and 1240.40 Å) behaves oddly
• Photospheric absorption of primary dominant (1350 – 1461 Å)
GHRS spectrum
• Emission features attributed to the disk used to calculate the
secondary’s radial velocity
• Calculated the average velocity of the peaks and use the result as
the kinematic tracer of the secondary
• Determined mass ratio q = 1.2 ± 0.6
• F-star mass 19 ± 14 M, B-star mass 16 ± 9 M
• Disk is opaque with inner edge irradiated and ionized
producing emission lines
• Cooler regions toward the outer edge of the disk are
detectable via absorption lines during primary eclipse
• Asymmetry suggests a hot spot from impact of accreting gas
• Possibly penetrates disk
Fig. 1. Dereddened spectrum of Epsilon Aurigae (thin line) compared to 2D
SEI Fe II wind model (thick line). Other prominent emission lines are marked.
Ake, T. (2006)
Fig. 2. Fe II profiles from the ground state (left) and excited levels (right). The
hatched areas show the location of four interstellar components seen in Na I
and K I (Hobbs 1974). The existence of blueshifted component (dashed thin
line) in excited levels indicates it arises from warmer circumstellar material
than the ISM. Ake. T. (2006)
FUSE spectrum
• Observed shortly after the predicted emergence of the
secondary object from behind the primary star (January 2001)
• Resolution ≈ 10,000 (?)
• Emission due to resonance scattering of photons in the
expanding wind of the supergiant or disk from an occulted hot
source
• NOT rotationally produced at the inner boundary of the disk as
per Sheffer & Lambert (1999)
• NO chromospheric emission lines may be attributed to the Fstar
• If present, Fe II fluorescent lines pumped by Ly α
FUSE spectrum
• Blueshifted absorption component in excited levels from
warm circumstellar material
• Blueshifted absorption component in ground state levels from
ISM (see Fig. 2)
• Compared dereddened Eps Aur spectrum with a 2D SEI Fe II
wind model (7000 K blackbody and interstellar H2 absorption)
(see Fig. 1)
• The emission lines cannot be used to infer the mass of a
rotating disk
• Velocity separation insufficient: wind origin may be F star, hot
central star or hot boundary layer in the disk
Results
1. UV continuum is variable out of eclipse - active phases
1.
2.
3.
Intrinsic to hot secondary (B-star)
Non-homogenous or intrinsically variable disk
Cepheid-like pulsations of the primary (F-star)
• Relate large amplitude changes in the UV excess to Cepheid-like
variations of the primary (Altner et al. 1986)
2. UV lines are variable out-of-eclipse
3. Eclipse is gray
1.
2.
3.
Opacity from electron scattering
Opacity from dust
Strong blended absorption lines from the eclipsing body affects
the position of the continuum especially in low-resolution
spectra (Boehm et al. 1984)
Next steps
• Hypothesis: Hα is a tracer for F-star chromospheric variability
•
•
•
•
•
Hα emission originates in F-star chromosphere
Compare Hα RV & EW time variations
Any correlation to magnitude? Or other stellar measurements?
How to explain polarization position angle variation with orbit?
Ca II H & K lines are very deep and polarization signal noisy. Can
we reduce noise in the line? If no/low emission, implications for
chromosphere?
Howell, S., Stencel, R. & Hoard, D. (2011)
Howell, S., Stencel, R. & Hoard, D. (2011)
Howell, S., Stencel, R. & Hoard, D. (2011)
NIST C II data
NIST ionization data
1 eV corresponds to about 11,600 K (NIST)
IUE publications
1. The partial phase of the eclipse of epsilon Aurigae
• Chapman, R. D. et al 1983ApJ...269L..17C
2. Epsilon Aurigae in eclipse. I - Ultraviolet spectroscopy during
ingress and totality
• Parthasarathy, M. et al 1983PASP...95.1012P
3. The eclipse of Epsilon Aurigae in the ultraviolet
• Boehm, C.; Ferluga, S.; Hack, M. 1984A&A...130..419B
4. Epsilon Aurigae
• Champman, R. 1985Ap+SS…110…177C
5. IUE observations of the 1982 - 84 eclipse of Epsilon Aurigae
• Ake, T. B. 1985NASCP2384...37A
IUE publications
6. High-dispersion spectroscopy of the eclipse of Epsilon
Aurigae at visible and ultraviolet wavelengths
• Ferluga & Hack 1985A+A_144_395F
7. Scattered light in the IUE spectra of Epsilon Aurigae
• Altner, B. et al 1986A&AS...65..199A
• Also mentions UVS
8. High-dispersion spectroscopy of the eclipse of Epsilon
Aurigae at visible and ultraviolet wavelengths
• Ferluga, S.; Hack, M. 1985A&A...144..395F
9. Ultraviolet flux variation of Epsilon Aurigae
• Kang, Young-Woon 1990JASS....7...57K
FUSE publication
1. Epsilon Aurigae – Is the Answer Blowin’ in the Wind?
• Ake, T. 2006ASPC..348..156A
GHRS publication
1. Intereclipse Spectroscopic Snapshot of ε Aurigae with the
Hubble Space Telescope
• Sheffer, Y. and Lambert, D. 1999PASP..111..829
COS publication
1. Hubble Space Telescope Ultraviolet Observations of Epsilon
Aurigae
• Howell, Steve B.; Stencel, R. E.; Hoard, D. W. (2011)
• http://adsabs.harvard.edu/abs/2011AAS...21725707H
Other publications
1. Flare Activity of epsilon Aurigae?
• Information Bulletin on Variable Stars, 2405, 1.
• Nha, I. –S. and Lee, S. J.
• http://adsabs.harvard.edu/abs/1983IBVS.2405....1N