1: %\documentclass[12pt,preprint]{aastex}
2: %\usepackage{emulateapj5}
3: \documentclass{emulateapj}
4: \newcommand{\vdag}{(v)^\dagger}
5: \newcommand{\myemail}{hwlee@sejong.ac.kr}
6: \slugcomment{To appear in ApJ}
7:
8: \shorttitle{Raman Scattered O~VI 6825 in V1016 Cyg and HM Sge}
9: \shortauthors{Lee and Kang}
10:
11: \begin{document}
12:
13: \title{Raman Scattered O~VI~$\lambda$~6825 and the Accretion Disk
14: Emission Model in the Symbiotic Stars V1016~Cygni and HM Sagittae}
15:
16: \author{Hee-Won Lee and Suna Kang}
17: \affil{Department of Astronomy and Space Science,
18: Astrophysical Research Center for the Structure and Evolution
19: of the Cosmos, Sejong University, Seoul, 143-747, Korea}
20:
21: \begin{abstract}
22:
23: We present the high resolution spectra of the D type symbiotic stars
24: V1016~Cygni and HM~Sagittae obtained with the Bohyunsan Optical
25: Echelle Spectrograph (BOES), and investigate
26: the double-peaked asymmetric profiles of the Raman scattered O~VI 6825.
27: This feature is formed through Raman scattering of O~VI 1032
28: by atomic hydrogen with small scattering cross section of
29: $\sim 10^{-23}{\rm\ cm^2}$, requiring a specific condition of coexistence
30: of a highly ionized emission nebula and a thick neutral region.
31: By adopting a wind accretion disk model,
32: we assume that the O~VI emission region is described by a Keplerian
33: thin disk. The Raman scattering occurs in a neutral region near the giant,
34: taking in the form of a slow stellar wind, part of which
35: is ionized by the strong UV radiation from the hot white dwarf.
36: Using a Monte Carlo technique, we compute the line profiles that are
37: modulated by the slow spherical stellar wind from the giant component
38: with the ionization front approximated by a hyperboloid.
39: In order to account for the asymmetry and the existence of a central dip
40: in the profiles, we add an O~VI resonance scattering region
41: between the hot white dwarf and the giant star which hinders the incidence
42: of slightly blue O~VI photons upon the H~I region.
43: Overall good fits to the observed data are obtained from our model,
44: which lends support to the accretion disk emission model in these objects.
45: The best fitting parameters for V1016~Cyg are $v_o=30{\rm\ km\ s^{-1}}$,
46: $v_\infty=11{\rm\ km\ s^{-1}}$, and $v_{c}=10{\rm\ km\ s^{-1}}$,
47: where $v_o$, $v_\infty$ and $v_{c}$ are the velocity of the outer disk
48: rim, the terminal velocity of the giant wind, and the velocity component
49: of the resonance scattering O~VI region along the binary axis,
50: respectively. Similar fitting parameters
51: $v_o=27{\rm\ km\ s^{-1}}$, $v_\infty=10{\rm\ km\ s^{-1}}$ and
52: $v_{c}=9{\rm\ km\ s^{-1}}$ are obtained for HM~Sge.
53: We also investigate the effect of a hot spot in a disk that is well known in
54: accretion disks in cataclysmic variables.
55: However, the introduction of a hot spot in our Keplerian disk
56: model failed to improve the overall profile fitting quality significantly.
57: Brief discussions about our profile analysis in relation to
58: bipolar morphology and accretion processes are presented.
59: \end{abstract}
60:
61: \keywords{binaries : symbiotic --- accretion, accretion disks ---
62: line : profiles --- scattering --- stars (individual V1016 Cyg, HM Sge) }
63:
64: \section{Introduction}
65:
66: Symbiotic stars are spectroscopically characterized by prominent emission lines
67: with TiO absorption bands that are typical of a giant. They are generally
68: known to be wide binary systems consisting of a hot white dwarf and a
69: giant star (e.g. Kenyon 1986). They are usually divided into
70: 'S' and 'D' type symbiotic stars, where 'D' type systems exhibit infrared
71: excess indicative of warm dust with temperature $T\sim 10^3{\rm\ K}$.
72: Many 'D' type symbiotic stars are believed to contain a Mira type
73: giant as cool component. Their orbital separation of the hot and
74: cool components may lie in the range of 10-100 AU, from which we expect
75: the orbital periods of order $10^2{\rm\ years}$ (Whitelock 1987).
76:
77: The giant component may lose mass
78: in the form of slow stellar wind, a significant fraction of which can be
79: photoionized by the hot white dwarf component. Taylor
80: \& Seaquist (1984) investigated the ionization structures in order to
81: interpret their radio observations of symbiotic stars.
82: Many prominent emission lines in symbiotic stars can
83: be excellent diagnostic tools to study the physical conditions
84: around the white dwarf component (e.g. M\"urset et al. 1991).
85:
86: It has been proposed by a number of researchers that some fraction
87: of wind material from the giant component may be gravitationally captured
88: by the hot component (e.g. Sokoloski
89: Bildsten \& Ho 2001). The mass loss rate from the giant component
90: may range from $10^{-4}-10^{-7}{\rm\ M_\odot\ yr^{-1}}$ depending
91: on the stellar evolution stage (e.g. Iben \& Tutukov 1996, Luthardt 1992).
92: Wind accretion in binary star systems was investigated
93: using the smoothed particle hydrodynamics (SPH) by Theuns
94: and Jorrissen (1993), who concluded that an accretion dis can be
95: formed in their isothermal flow models.
96: Mastrodemos \& Morris (1998) also adopted an SPH method
97: to show that an accretion disk can be formed around the white dwarf
98: component. There are only limited studies about
99: the existence of accretion disks and the insufficient numerical resolution
100: prevents one from deducing the physical properties of an accretion disk
101: in symbiotic stars. Symbiotic stars being considered
102: as a candidate for type Ia supernova progenitors, the mass transfer
103: process in symbiotic stars can also be a very important subject in other
104: branches of astronomy.
105:
106: About a half of symbiotic stars are known to exhibit very broad and
107: mysterious emission features around 6825 \AA\ and 7088 \AA\ that were
108: suspected to originate from highly ionized species (Allen 1980).
109: These features were finally identified by Schmid (1989), who proposed the
110: Raman scattering origin. According to him, the Raman
111: scattering process starts with an O~VI 1032 line photon incident
112: upon a hydrogen atom in the ground state, which can be converted
113: into an optical 6825 photon
114: as a result of an inelastic scattering off the hydrogen atom
115: that is left in its excited $2s$ state. From an incident O~VI 1038
116: line photon, we obtain a Raman photon at around 7088 \AA.
117: The Raman scattering nature of these features is strongly supported by
118: many observational studies including spectropolarimetric observations by
119: Harries \& Howarth (1996) and
120: contemporaneous UV and optical observations by Birriel et al. (2000).
121:
122: The wavelength $\lambda_o$ of the inelastically
123: scattered radiation is related with the incident wavelength $\lambda_i$
124: and $\lambda_{Ly\alpha}$ of hydrogen Ly$\alpha$
125: by
126: \begin{equation}
127: {\lambda_o}^{-1}={\lambda_i}^{-1}-\lambda_\alpha^{-1}.
128: \end{equation}
129: Differentiation of this equation leads to
130: \begin{equation}
131: {\Delta\lambda_o\over\lambda_o}=
132: \left({\lambda_o\over\lambda_i}\right){\Delta\lambda_i
133: \over\lambda_i}.
134: \label{wavelength}
135: \end{equation}
136:
137: In the case of Raman scattering involving O~VI 1032, $\lambda_o/\lambda_i
138: =6.6$.
139: This provides two important characteristics in the profiles of Raman
140: scattered O~VI 6825
141: that they are broader than their parent emission lines by a factor
142: of $\lambda_o/\lambda_i$ and that they are mainly determined
143: by the relative motion between the neutral scatterers and line photon
144: emitters. Therefore, the profiles are little affected by the direction of the
145: observer's line of sight. This is due to the inelasticity
146: of Raman scattering and noted to be one of the most important spectroscopic
147: properties retained by Raman scattered features (e.g. Nussbaumer, Schmid \&
148: Vogel 1989, Schmid 1989). In this respect, Raman scattered lines provide
149: a very interesting opportunity to study the accretion process involving
150: a slow stellar wind, which is quite unique to symbiotic stars.
151:
152: Thus far, Raman scattered lines
153: are found for O~VI 1032, 1038 resonance doublets and He II emission lines
154: near H~I Lyman series in many symbiotic stars and in a few young
155: planetary nebulae (e.g. van Groningen 1993,
156: P\'equignot et al. 1997, Groves et al. 2002, Lee et al. 2003, Birriel 2004,
157: Zhang et al. 2005, Lee et al. 2006b). Raman scattered O~VI 6825 from O~VI 1032
158: is particularly strong in a number of symbiotic stars, for which a
159: refined profile analysis is feasible. Furthermore,
160: Raman scattered features consist of purely scattered photons, not mixed
161: with the direct emission, which makes them characterized by strong linear
162: polarization.
163:
164: Fairly intensive spectropolarimetric studies of Raman scattered O~VI lines
165: were carried out by Harries \& Howarth (1996). They found that
166: the Raman scattered O~VI 6825 and 7088 show double or
167: triple peak profiles and that they are strongly polarized with
168: complicated structures.
169: Schmid (1996) used a Monte Carlo method to calculate
170: the profiles and polarization of these scattered features formed
171: in a slowly expanding neutral region that mimics the stellar wind around
172: the giant component. A similar study with a slightly different velocity
173: law of the slow stellar wind was performed by Lee \& Lee (1997b).
174:
175: In these studies, no kinematic structure is assumed for the emission
176: region, which is treated effectively as point-like
177: without any complicated internal structures.
178: Instead, the stronger red part of Raman scattered O~VI 6825 is
179: mainly attributed to scattering at the receding part of the
180: giant stellar wind that has a bigger geometric covering
181: of the O~VI emission region than the approaching part does.
182: However, in these models, one has to invoke
183: a rather high velocity scale $\ge 50{\rm\ km\ s^{-1}}$ of the stellar wind
184: from the giant in order to explain the observed Raman profiles.
185:
186: Lee \& Park (1999) adopted an accretion disk emission model and
187: proposed the origin of the multiple peak profiles with strong linear
188: polarization. In their picture, the accretion flow
189: can be affected significantly by the binary orbital motion, which results
190: in an asymmetric matter distribution around the white dwarf causing
191: the stronger red part of the Raman features.
192: However, their study lacked detailed calculations that can be directly
193: compared with the observed profiles.
194:
195: Both V1016~Cyg and HM~Sge are classified as symbiotic novae or slow novae.
196: V1016~Cyg underwent a slow nova-like outburst in 1965, which may be
197: interpreted to be a thermonuclear runaway on the surface of the white dwarf
198: (Fitzgerald et al. 1966). A similar nova-like optical outburst occurred
199: in HM Sge in 1975 (Dokuchaeva 1976). These two symbiotic stars are known
200: to exhibit fairly strong Raman scattered O~VI 6825 with
201: characteristic double peak profiles. In particular, the appearance of
202: Raman scattered 6825 in HM~Sge was reported by Schmid et al. (2000)
203: in their spectra obtained with the William Herschel Telescope (WHT) in 1998.
204:
205: Both V1016~Cyg and HM~Sge are D-type symbiotic stars,
206: containing a Mira type giant as cool component.
207: Thus far, no detailed information is available about
208: the orbital parameters of V1016~Cyg and HM~Sge.
209: Schild \& Schmid (1996) measured the rotation of position
210: angle in the polarized Raman 6825 with an estimation of the
211: orbital period of $\sim 80{\rm\ yr}$ in V1016~Cyg. However, Schmid (1998)
212: noted that this estimate may be severely affected by the change of the nebular
213: structure in this system and the orbital period may exceed 100 years
214: (e.g. Schmid 2001). Using the {\it Hubble Space Telescope (HST)} data,
215: Brocksopp et al. (2002)
216: measured the angular separation of $42.4{\rm\ mas}$ between the white dwarf
217: and the Mira of V1016~Cyg, proposing the physical binary separation of
218: $\sim 84{\rm\ AU}$ with their adopted distance of $2{\rm\ kpc}$.
219: According to the {\it HST} study of HM~Sge by Eyres et al. (2001),
220: the separation of the Mira
221: and the white dwarf is $\sim 50{\rm\ AU}$ with the adopted distance of
222: $\sim 1.2 {\rm\ kpc}$.
223:
224:
225: In this paper, we present our spectra
226: of these Raman features in the symbiotic stars V1016~Cyg and HM Sge
227: secured with the Bohyunsan Optical
228: Echelle Spectrograph (BOES),
229: and make quantitative comparisons with the profiles
230: computed by a Monte Carlo technique adopting an accretion disk emission
231: model.
232:
233:
234: \section{Observation and Data}
235:
236: We observed V1016~Cyg and HM~Sge on the nights of 2005 November 7 through 9
237: with the Bohyunsan Optical Echelle Spectrograph installed on the
238: 1.8~m telescope at Mt. Bohyun. The detector was a 2k$\times$4k
239: E2V CCD with pixel size $15{\rm\ \mu m}\times 15 {\rm\ \mu m}$, in which
240: about 80 spectral orders covering $\sim 3700{\rm\ \AA}$ to
241: $\sim 10,000{\rm\ \AA}$ are recorded in a single exposure.
242: The 300~$\mu$ optical fiber was used
243: to yield the spectroscopic resolving power
244: of $R=30,000$ with an aperture field of 6.4 arcseconds.
245: A more detailed description of BOES can be found in
246: Lee et al. (2006b) and in Kim et al. (2002).
247: The exposure times are from 7000~s and 3600~s for V1016~Cyg and HM~Sge,
248: respectively. The data have been
249: reduced following standard procedures using the IRAF packages.
250:
251: Parts of our spectrum around 6825~\AA\ of V1016~Cyg and HM Sge are
252: shown in Fig.~1.
253: In the figure, the line flux is normalized by the local continuum
254: in order to focus on the profile shape in the current work.
255: The two profiles are overall similar to each other.
256: The 6825 feature of HM~Sge is much weaker and therefore of poorer quality
257: than that of V1016~Cyg, which is indicated by the larger scatter
258: in the data. In both systems, the red part is relatively stronger
259: than the blue part, which is noted by many previous researchers
260: (e.g. Harries \& Howarth 1996, Schmid et al. 1999).
261:
262: There exists a narrow emission line at around 6820 \AA\ in HM~Sge.
263: The presence of this narrow emission feature
264: in HM~Sge was also noted in the spectropolarimetric studies by
265: Schmid et al. (2000). They concluded that this emission line is
266: not related with Raman scattering on the ground that it is unpolarized
267: and narrow. We ignore this feature in the subsequent analysis, by
268: subtracting it by a single Gaussian $f(\lambda)=f_0 \exp[(\lambda
269: -\lambda_c)^2/\Delta\lambda^2]$ with $f_0=0.43,\
270: \lambda_c=6820.09 {\rm\ \AA}$ and $\Delta\lambda = 0.46{\rm\ \AA}$.
271:
272: In the case of V1016~Cyg, the local maxima are found at
273: 6823.5 \AA\ and 6830.8 \AA. The separation of these two peaks
274: corresponds to $48{\rm\ km\ s^{-1}}$ in the velocity space
275: of O~VI 1032 emitters.
276: The FWZI (full width at zero intensities) is $166{\rm\ km\ s^{-1}}$.
277: It is noted that there is an asymmetrically extended red wing
278: that may end at around 6842.7 \AA.
279:
280:
281: The peaks for HM~Sge are located at 6823.3 \AA\ and 6829.4 \AA,
282: for which the separation in velocity space of O~VI is $41{\rm\ km\ s^{-1}}$.
283: This is significantly smaller than that for V1016~Cyg.
284: The FWZI is $161{\rm\ km\ s^{-1}}$, which is comparable to that
285: of V1016~Cyg. In contrast with V1016~Cyg, it appears that there is
286: no (or significantly weak) extended red wing in HM~Sge. However,
287: this should be taken with the
288: caution that the data quality of HM~Sge is much poorer than V1016~Cyg.
289:
290: The relative kinematics between
291: the O~VI emission region and neutral H~I scattering region in both
292: V1016~Cyg and HM~Sge is
293: characterized by velocity scales ranging from
294: $20{\rm\ km\ s^{-1}}$ to $80{\rm\ km\ s^{-1}}$. If the FWZI of the
295: profiles is attributed to the kinematics of the giant wind as
296: in the models of Schmid (1996) and Lee \& Lee (1997b), then the wind terminal
297: speed of $80{\rm\ km\ s^{-1}}$ should be required, which is much faster
298: than the escape velocity of a giant star.
299:
300: \begin{figure}
301: %\epsfig{file = obs_boes.ps, scale=0.67, angle=270}
302: %\epsfig{file = f1.eps}
303: \includegraphics[angle=0,scale=.60]{f1.eps}
304: \caption{Parts of spectra around 6825~\AA\ of the symbiotic
305: stars V1016~Cyg and
306: HM~Sge obtained with the Bohyunsan Optical Echelle Spectrograph (BOES).
307: The vertical scale is normalized by the local continuum.
308: The peak separation is $\sim 48{\rm\ km\ s^{-1}}$ for V1016~Cyg
309: and $\sim 41 {\rm\ km\ s^{-1}}$ for HM~Sge. The full widths at zero
310: intensities are $166{\rm\ km\ s^{-1}}$ and $161{\rm\ km\ s^{-1}}$
311: for V1016~Cyg and HM~Sge, respectively.
312: }
313: \end{figure}
314:
315: \section{Monte Carlo Line Profile Analysis}
316: \subsection{Accretion Disk Model}
317:
318: In this work, we assume that the wind accretion occurs in both V1016~Cyg
319: and HM Sge to form an accretion disk around each white dwarf component.
320: According to Model 3 of the SPH studies by Mastrodemos \& Morris (1998),
321: an accretion disk with size ranging 0.4 - 0.8 AU is formed
322: in a binary system containing a white dwarf, where the binary orbital
323: period is 18 years. We consider that the O~VI emission region is limited
324: to a part of the whole accretion flow, which is approximated to be Keplerian.
325: For simplicity, we assume that
326: the O~VI emission region is confined to an annular region
327: characterized by the inner radius $R_i$
328: and the outer radius $R_o$.
329:
330:
331: In this work we assume that each white dwarf component has a mass
332: of $0.7 {\rm\ M_\odot}$. Noting that the Raman scattered O~VI
333: 6825 in both V1016 Cyg and HM Sge extends about $\Delta v\sim 160{\rm\ km\
334: s^{-1}}$, we infer that the O~VI emission region extends from $\sim 0.05$ AU
335: to less than 1 AU.
336:
337: %With these ingredients we prepare a characteristic double peak profile
338: %of O~VI 1032, we consider the disk component with a constant emissivity
339: %$\epsilon_0$ for simplicity. In our Monte Carlo code,
340: %when a given random number $r_p$ in the interval $(0,1)$ is given,
341: %a photon is generated at $r$ by
342: %\begin{equation}
343: %{r\over R_o} =
344: %\left[\left( {R_i\over R_o} \right)^{2}
345: %+\left( 1-\left\{ R_i\over R_o \right\}^{2} \right)r_p
346: %\right]^{1/2}.
347: %\end{equation}
348:
349: With these ingredients we prepare a characteristic double peak profile
350: of O~VI 1032, we first consider the disk component with the emissivity
351: $\epsilon({\bf r})$ that depends only on $r$ through a power law
352: \begin{equation}
353: \epsilon({\bf r}) = Ar^{\alpha}, \ R_i<r<R_o
354: \label{eqalpha}
355: \end{equation}
356: where $A$ is a constant. This functional dependency of emissivity is
357: achieved in our Monte Carlo code by specifying the location $r$
358: of the initial photon from a uniform random number $r_p$ in the interval
359: $(0,1)$ with the prescription
360: \begin{equation}
361: {r\over R_o} =
362: \left[\left( {R_i\over R_o} \right)^{\alpha+2}
363: +\left( 1-\left\{ R_i\over R_o \right\}^{\alpha+2} \right)r_p
364: \right]^{1/(\alpha+2)}.
365: \end{equation}
366:
367:
368: With another random number $r_\phi$ in the interval $(0, 2\pi)$, we
369: locate the point of generation of the incident photon $P(r,\phi=2\pi
370: r_\phi)$. Then the photon is assigned a Doppler factor $DF$ along
371: the propagation direction
372: \begin{equation}
373: DF = {\bf k}_i\cdot {\bf v}(P)/c,
374: \end{equation}
375: where ${\bf v}(P)=\sqrt{GM_{WD}\over r}(-\sin\phi{\bf\hat
376: x}+\cos\phi{\bf\hat y})$ is the local Keplerian velocity at $P$.
377:
378: According to Eq.~(\ref{wavelength}), the observed wavelength of Raman
379: scattered O~VI is determined
380: by the relative motion between the O~VI emitter and the H~I scatterer
381: and quite insensitive to the observer's line of sight.
382: Furthermore, the orbital parameters of V1016~Cyg and HM~Sge
383: are only poorly known.
384: In view of these facts, in our Monte Carlo code, we collect all the photons
385: scattered inelastically off hydrogen atoms around the giant
386: component. However, a more refined study should include the effect
387: of the observer's line of sight with full care.
388:
389:
390: \subsection{Velocity Modulations from a Slow Stellar Wind}
391:
392: The direct O~VI 1032 line photons are scattered in a neutral scattering
393: region, which may be expanding with a velocity $\sim 10{\rm\ km\
394: s^{-1}}$. Schmid (1996) investigated the line profile formation of
395: Raman scattered O~VI in an expanding slow stellar wind using a Monte
396: Carlo technique (see also Harries \& Howarth 1997).
397: A similar study was also conducted by Lee \& Lee
398: (1997b) using a Monte Carlo method with a slight different kinematical
399: velocity law for the slow stellar wind.
400:
401: Adopting also a Monte Carlo technique, we extend the work of Lee \&
402: Lee (1997b) to incorporate the accretion
403: disk emission model considered in the previous section. The velocity
404: of the slow and spherical stellar wind ${\bf v}_H({\bf r})$ at the position
405: ${\bf r}$ is given by
406: \begin{equation}
407: {\bf v}_H({\bf r}) =v_\infty (1-R_*/r)^\beta {\bf\hat r},
408: \label{eqgwind}
409: \end{equation}
410: where $v_\infty$ is the terminal wind velocity and the origin of the
411: coordinate system coincides with the center of the giant. For
412: simplicity of the current investigation, we set the power $\beta=1$ of the
413: velocity law. We introduce a new dimensionless vector defined by
414: \begin{equation}
415: {\vec \rho}={\bf r}/R_*.
416: \end{equation}
417: Assuming a constant mass loss rate $\dot M$, the density
418: $n{\bf r})$ is described by
419: \begin{equation}
420: n({\bf r}) = n_0\rho^{-2}(1-\rho^{-1})^{-1},
421: \label{nr}
422: \end{equation}
423: where $n_0=\dot M/4\pi\mu R_*^2 m_pv_\infty$ with $m_p$ and $\mu$
424: being the mass of a proton and the mean molecular weight,
425: respectively.
426:
427: In the Monte Carlo calculation, it is essential to locate the next
428: scattering position ${\vec\rho}_2$ for a given starting
429: position ${\vec\rho}_1$ in the direction ${\bf\hat k}_i$, the unit
430: wavevector of the photon under consideration.
431: In order to accomplish this, an optical depth $\tau_{12}$ is
432: generated using a uniform random number $r_{tau}$ between 0 and 1 by
433: \begin{equation}
434: \tau_{12}=-\ln r_{tau}.
435: \end{equation}
436: In terms of the physical distance $s_{12}$ corresponding to $\tau_{12}$
437: from the starting position ${\vec\rho}_1$ along the direction
438: ${\bf\hat k}_i$, the new scattering position is expressed as
439: \begin{equation}
440: {\vec\rho}_2 = {\vec\rho}_1 + s_{12}{\bf\hat k}_i.
441: \label{s12}
442: \end{equation}
443: The new scattering position $\vec\rho_2$ is related with $\tau_{12}$ by
444: \begin{equation}
445: \tau_{12} = \int_{{\vec\rho}_1}^{{\vec\rho_2}}\ n({\bf r})
446: \sigma_{tot}\ dl,
447: \label{tau12d}
448: \end{equation}
449: where $\sigma_{tot}$ is the sum of the cross sections for Rayleigh and
450: Raman scattering. Lee \& Lee (1997a) proposed the value
451: $\sigma_{tot}=42\sigma_T=2.8\times 10^{-23}{\rm\ cm^2}$
452: with $\sigma_T=6.6\times 10^{-25}{\rm\ cm^2}$ being the Thomson
453: scattering cross section, which we adopt in this work.
454:
455: We introduce the representative optical depth
456: $\tau_0$ defined by
457: \begin{equation}
458: \tau_0 = n_0 R_*\sigma_{tot} =8.8 \dot M_{-6}\ R_{*13}\ v_{\infty10},
459: \end{equation}
460: where $\dot M_{-6}=\dot M/(10^{-6}{\rm\ M_\odot\ yr^{-1}})$,
461: $R_{*13}=R_*/(10^{13}{\rm\ cm})$ and $v_{\infty10}=v_{\infty}/(10{\rm\
462: km\ s^{-1}})$.
463: The exact choice of $\tau_0$ and $\sigma_{tot}$
464: affects the ionization
465: structure and hence the exact profile shape. However, the ionization structure
466: is also dependent on the UV radiation from the hot component. In view of this,
467: we set $\tau_0=2.5$ in our simulations, in order to reduce the parameter
468: space of investigation.
469:
470: We also define $b$ as the impact parameter of the photon path with respect to
471: the giant center, and the parameter $s$ measures the physical distance
472: along the photon path. For the sake of convenience, we set $s=0$
473: at the foot of the perpendicular from the giant center to the photon
474: path, so that
475: \begin{equation}
476: s=\pm\sqrt{\rho^2 - b^2},
477: \label{impact}
478: \end{equation}
479: where negative $s$ is obtained when ${\bf\hat k}_i\cdot {\vec\rho}_1
480: <0$. In terms of these parameters, the integral (\ref{tau12d}) can be
481: rewritten as
482: \begin{equation}
483: \tau_{12} =\int_{s_1}^{s_2} n(s)\ \sigma_{tot}\ ds
484: =\tau_0\int_{\rho_1}^{\rho_2} {d\rho\over\sqrt{\rho^2-b^2}(\rho-1)}.
485: \label{tau12}
486: \end{equation}
487: In terms of a new function $u(\rho)$ defined by
488: \begin{equation}
489: u(\rho)={1\over\sqrt{|b^2-1|}}[\rho-1+\sqrt{\rho^2-b^2}],
490: \label{urho}
491: \end{equation}
492: the integral in Eq.(\ref{tau12}) can be explicitly expressed as
493: \begin{equation}
494: \tau_{12}={2\tau_0\over\sqrt{|b^2-1|}}[\tan^{-1}u(\rho_2)
495: -\tan^{-1|}u(\rho_1)].
496: \end{equation}
497: Using the inverse relation of Eq.(\ref{urho})
498: \begin{equation}
499: \rho=b\cosh \ln [(\sqrt{b^2-1}u+1)/b]
500: \end{equation}
501: this relation is directly inverted to obtain $\rho_{2}$
502: in terms of $\tau_{12}$ and $\rho_1$.
503:
504: A more complicated situation is obtained when $b>1,
505: {\bf\hat k}_i\cdot{\bf r}_1<0$, in which case we consider
506: $\tau_b$, the optical depth to the point of impact, defined by
507: \begin{equation}
508: \tau_b %= \int_{0}^{s_1} \ n(s)\ \sigma_{tot}\ ds
509: = {2\tau_0\over\sqrt{b^2-1}}
510: \left[\tan^{-1}u_1-\tan^{-1}\left({ b-1 \over b+1 }\right)^{1/2}\right].
511: \end{equation}
512: If $\tau_{12}>\tau_b$, we consider
513: \begin{equation}
514: \tau_{12}=\tau_b+\int_0^{s_2} n(\rho)\ \sigma_{tot}\ ds,
515: \end{equation}
516: from which we obtain
517: \begin{equation}
518: u(\rho_2)=\tan{ \sqrt{b^2-1}(\tau_{12}-\tau_b)\over 2\tau_0}.
519: \end{equation}
520:
521: With $\rho_2$, $s_{12}$ is determined, and the location of
522: the next scattering position ${\vec \rho}_2$ is then given
523: by Eq.(\ref{s12}). With a branching
524: ratio $b_{Ram}\sim 0.15$, the scattering process is Raman, in which case
525: we assume that the Raman photon escapes from the region without further
526: interaction. If the scattering process is Rayleigh, then the photon
527: propagates along the new direction ${\bf\hat k}_o$, which is chosen
528: from the Rayleigh phase function (e.g. Schmid 1996, Lee \& Lee 1997a).
529:
530:
531: \subsection{Ionization Fronts}
532: \begin{figure}
533: %\epsfig{file=stromgren.ps, angle=270, scale = 0.67}
534: %\epsfig{file=f2.eps, angle=270, scale = 0.67}
535: \plotone{f2.eps}
536: \caption{Ionization structure in a symbiotic star
537: for a range of ionizing luminosities
538: represented by the parameter $X$ defined in Eq.~(\ref{ionx}).
539: The dots at the origin and
540: at the point $(80,0)$ stand for a mass losing giant and a hot white
541: dwarf. The volume of the H~I region increases as $X$ decreases.
542: The ionization front corresponding to $X=0.02$ is approximated by
543: a hyperbola given by $y^2=6.5(x-40)^2-400$.}
544: \label{ifront}
545: \end{figure}
546:
547: The part of the stellar wind from the giant facing the hot white
548: dwarf will be ionized by strong far UV radiation. Taylor \& Seaquist
549: (1984) investigated the ionization structures that
550: depend the density and the ionizing luminosity. They introduced a
551: a parameter $X$ defined by
552: \begin{equation}
553: X = {4\pi\mu^2m_p^2\over\alpha} aL_{ph} (v_\infty/\dot M)^2.
554: \label{ionx}
555: \end{equation}
556: Here, $\alpha$ is the recombination coefficient,
557: $L_{ph}$ is the luminosity of ionizing radiation, and
558: $a$ is the photoionization cross section.
559:
560: Then, the ionization front is described by the equation
561: \begin{equation}
562: X=f(u,\theta) \equiv\int_0^u n({\bf r})^2 s^2 ds
563: \end{equation}
564: where $s$ is the distance measured from the white dwarf. Taylor \&
565: Seaquist (1984) provided the closed form of $f(u,\theta)$
566: using the density law $n({\bf r})\propto r^{-2}$, which is slightly
567: different from the one given by Eq.~(\ref{nr}) adopted in this work.
568: Although the description of Taylor \& Seaquist (1984) is also
569: approximately valid for the current work, we present the detailed
570: ionization structure in a closed form in this subsection
571: for the sake of completeness.
572:
573: With the density law given by Eq.~(\ref{nr}) the necessary integral to
574: be carried out is
575: \begin{equation}
576: I=\int{s^2 ds\over \rho^2(\rho-1)^2}=
577: \int{(\rho_i\cos\phi\pm\sqrt{\rho^2-b^2})^2 d\rho
578: \over \rho(\rho-1)^2 \sqrt{\rho^2-b^2}},
579: \end{equation}
580: where $\rho$ and $s$ are related by the impact parameter $b$ as in
581: Eq.~(\ref{impact}).
582:
583: For $b=\epsilon^{-1}>1$, an explicit expression
584: of the integral $I$ is given by
585: \begin{eqnarray}
586: I &=& \rho_i\cos2\phi\csc\phi\tan^{-1}\sqrt{(\epsilon\rho)^2-1}
587: +{(\epsilon^2-\csc^2\phi)\sqrt{\epsilon^2\rho^2-1}
588: \over \epsilon(\rho-1)(1-\epsilon^2)}
589: \nonumber \\
590: &+&
591: {4\epsilon^2\cot^2\phi-2\epsilon^2+2\csc^2\phi \over
592: \epsilon(1-\epsilon^2)^{3/2}}
593: \tan^{-1}\left[ \left({1+\epsilon\over 1-\epsilon}\right)
594: \left({\epsilon\rho-1\over\epsilon\rho+1}\right) \right]^{1/2}
595: \nonumber \\
596: & \pm&
597: 2\rho_i\cos\phi
598: \left(\ln{\rho\over\rho-1}
599: -{1\over\rho-1}\right).
600: \label{eqifront1}
601: \end{eqnarray}
602: A more detailed calculation is given in the appendix.
603:
604:
605: In Fig.~\ref{ifront}, we show the ionization fronts for various values
606: of $X$. The horizontal axis coincides with the line connecting
607: the two stars. The dot at the origin stand for the mass losing giant.
608: We place the white dwarf at the point $(80,0)$, considering the
609: binary separation $\sim 80{\rm\ AU}$ proposed by Brocksopp et al. (2004).
610: The ionization front corresponding to $X=0.02$ is fitted by a hyperbola
611: given by $y^2=6.5(x-40)^2-400$. The angle between the horizontal axis
612: and an asymptote is $69^\circ$. This is somewhat larger than
613: the value proposed for V1016~Cyg by Jung \& Lee (2004). Nevertheless,
614: we perform Monte Carlo calculations using this value, because the exact
615: value of the half opening angle affects only the overall flux of the Raman
616: 6825 feature and the profile shape is less sensitive.
617:
618:
619: \section{Results}
620:
621: \subsection{Pure Keplerian Disk}
622:
623: \begin{figure}
624: %\epsfig{file = purekep.ps, scale = 0.67, angle = 270}
625: %\epsfig{file = f3.eps, scale = 0.67, angle = 270}
626: \plotone{f3.eps}
627: \caption{
628: Angle-averaged Monte Carlo line profiles of Raman scattered
629: 6825 from Keplerian disks for various $R_i/R_o$ with $R_i$ and $R_o$
630: being the inner radius and the outer radius of the disk.
631: A stationary scattering H~I region is assumed.
632: Main peaks are found at $\pm v_o$, the velocity
633: at the outer rim of the disk, which is set to $30{\rm\ km\ s^{-1}}$.
634: }
635: \label{purekep}
636: \end{figure}
637:
638: In Fig.~\ref{purekep}, we show the profiles of Raman scattered 6825
639: formed in a stationary H~I region relative to a Keplerian disk.
640: The H~I region is modeled as the hyperboloid given in Fig.~\ref{ifront}
641: of which the asymptotic cone has an half opening angle $\theta_o=69^\circ$.
642: In the figure, we fix the velocity $v_o=30{\rm\ km\ s^{-1}}$
643: at the outer edge $r=R_o$, and varied the inner radius $R_i$ from
644: $0.01R_o$ to $0.9R_o$. In the absence of a bright spot, the emission
645: profile is symmetric with respect to the line center.
646:
647: It is quite clear that the main peaks correspond to the velocity $v_o$
648: at the outer rim of the disk, which dominantly contributes to the
649: whole line flux. The emission region is almost ring-like in
650: the case $R_i=0.9R_o$, where the profile has a sharp edge at $\pm v_o$.
651: With decreasing $R_i$ the profile broadens around the sharp edges
652: at $\pm v_o$. However, when $R_i \le 0.1R_o$, the profiles are effectively
653: similar, because the inner disk is simply physically too small
654: to contribute to the overall profile significantly.
655: In view of this, we fix the ratio of the inner and outer radii to
656: be $R_i/R_o=0.1$ from now on.
657:
658: \begin{figure}
659: %\epsfig{file=alpha.ps, scale=0.67, angle = 270}
660: %\epsfig{file=f4.eps, scale=0.67, angle = 270}
661: %\epsscale{0.8}
662: \plotone{f4.eps}
663: \caption{Line profiles from a Keplerian disk with the emissivity
664: given by $\epsilon(r) = Ar^\alpha$, where $\alpha=\pm1$ and $\alpha=0$.
665: Solid line shows a profile for $\alpha=0$, dotted line for $\alpha=-1$
666: and long dashed line for $\alpha=1$. Far wings,
667: contributed from the inner disk region, get stronger
668: as $\alpha$ decreases.
669: }
670: \label{alpha}
671: \end{figure}
672:
673: In Fig.~\ref{alpha}, we show three line profiles for $\alpha=-1, 0$ and 1,
674: where $\alpha$ is defined in Eq.~\ref{eqalpha}. In this figure, $R_i=0.1R_o$
675: is fixed in order to focus on the dependence of profiles on the functional
676: form of the emissivity. With negative $\alpha$, the contribution from
677: the inner disk region becomes larger, which leads to a broader
678: profile with enhanced wings.
679: However, the locations of the main peaks are fixed at velocities
680: $v=\pm v_o$, yielding qualitatively overall similar profiles
681: irrespective of the detailed values of $\alpha$.
682: From this consideration, we fix $\alpha=0$ in the remainder of this paper.
683:
684: \subsection{Profile Modulation by the Giant Wind}
685:
686: \begin{figure}
687: %\epsfig{file = gwind.ps, scale = 0.67, angle = 270}
688: %\epsfig{file = f5.eps, scale = 0.67, angle = 270}
689: \plotone{f5.eps}
690: \caption{Angle-averaged line profiles of Raman scattered 6825
691: from a Keplerian disk that are formed in a spherically
692: expanding H~I region around the Mira component characterized
693: by the terminal wind velocity $v_\infty$. The H~I region is partially
694: ionized with the ionization front given by the hyperboloid
695: shown in Fig.~\ref{ifront}. With smaller $v_\infty$ the double-peak
696: structure is more conspicuous, and it is completely erased when
697: $v_\infty \ge 15{\rm\ km\ s^{-1}}$.
698: }
699: \label{giantwind}
700: \end{figure}
701:
702: In Fig.~\ref{giantwind}, O~VI line photons are incident on the
703: stellar wind around the giant that is expanding according to
704: Eq.~(\ref{eqgwind}).
705: In this figure, the wind terminal velocity $v_\infty$ has a range
706: from $5{\rm\ km\ s^{-1}}$ to $15{\rm\ km\ s^{-1}}$.
707: Here, we also fix the inner radius $R_i=0.1R_o$, and other parameters
708: to be the same as those considered in Fig.~\ref{purekep}. Therefore
709: Fig.~\ref{purekep} is obtained when we set $v_\infty = 0$.
710:
711:
712: As is seen clearly in the figure, the sharp edges near $v=\pm v_o$
713: are smoothed by the expanding motion of the H~I region.
714: When $v_\infty$ exceeds $13{\rm\ km\ s^{-1}}$
715: the double peak structure can hardly be noticed. From this, it is
716: concluded that the giant stellar wind with terminal velocity $<
717: 13{\rm\ km\ s^{-1}}$ is consistent with the double peak profiles
718: observed in V1016~Cyg and HM~Sge.
719:
720: However, the profile modulation from other sources such as
721: turbulence in the emission region may give rise to similar
722: profiles shown in Fig.~\ref{giantwind}. We do not consider this
723: possibility in the current work.
724:
725: \subsection{O~VI Resonance Scattering Region}
726:
727: \begin{figure}
728: %\epsfig{file=bestfit.ps, scale = 0.67, angle = 270}
729: %\epsfig{file=f6.eps, scale = 0.67, angle = 270}
730: \plotone{f6.eps}
731: \caption{Best fitting profiles from our Monte Carlo
732: calculations and the BOES spectra of V1016~Cyg and HM~Sge. Dots show
733: the BOES spectra and solid lines represent our Monte Carlo results.
734: The fitting parameters for V1016~Cyg are $v_o=30{\rm\ km\ s^{-1}},
735: v_\infty=11{\rm\ km\ s^{-1}}, v_c=-10{\rm\ km\ s^{-1}}$ and
736: $\Delta v_c = 9{\rm\ km\ s^{-1}}$.
737: For HM~Sge they are $v_o=26{\rm\ km\ s^{-1}},
738: v_\infty=10{\rm\ km\ s^{-1}}, v_c=-7{\rm\ km\ s^{-1}}$ and
739: $\Delta v_c = 6{\rm\ km\ s^{-1}}$.
740: }
741: \label{bestfit}
742: \end{figure}
743:
744: It is clear that the redward asymmetry exhibited in the Raman 6825 features
745: of V1016~Cyg and HM~Sge can not be accounted for by a pure Keplerian disk
746: emission model. It is quite uncertain whether we are seeing an excess in
747: red emission or a deficit in the blue part, or possibly both. An excess
748: red emission may correspond to a hot spot in an accretion disk which will
749: be discussed in the following subsection and in this subsection we consider
750: the first possibility.
751:
752: In order to account for the stronger red part in the observed profiles,
753: another component need to be added in the previous model considered
754: in Fig.~\ref{giantwind}. Therefore, we introduce an O~VI region between
755: the white dwarf and the giant. This additional O~VI region is assumed
756: to have a very low density so that the O~VI emission from this region is
757: negligible compared to the Keplerian emission disk. Nevertheless,
758: the line center optical depth exceeds unity in a velocity space
759: centered at $v_c<0$. Schmid et al. (1999) investigated O~VI 1032 profiles
760: and compared with those of non-resonant O~VI lines to conclude that
761: the line transfer effect due to resonance scattering of O~VI 1032
762: is present in symbiotic stars including V1016~Cyg.
763:
764: More specifically, we assume that the scattering optical depth
765: by the hot wind component is given by a Gaussian function
766: \begin{equation}
767: \tau_w = \exp[(v-v_c)^2/\Delta v_c^2].
768: \label{eqhotwind}
769: \end{equation}
770: This prescription hinders the incidence of those line photons
771: with the Doppler factor near $v_c/c$ upon the scattering region.
772: However, the velocity modulation due to the giant wind fills
773: the gap significantly, allowing one to obtain a double-peak profile
774: with some suppresion in the red part.
775:
776: Fig.~\ref{bestfit} shows the best fitting profiles
777: to out BOES spectra of V1016~Cyg
778: and HM~Sge. For V1016~Cyg, the velocity $v_o$ at the outer rim
779: of the accretion disk is $v_o=30{\rm\ km\ s^{-1}}$ and the
780: terminal velocity of the giant wind $v_\infty = 11{\rm\ km\ s^{-1}}$.
781: The corresponding values for HM~Sge are
782: $v_o = 26{\rm\ km\ s^{-1}}, v_\infty = 10{\rm\ km\ s^{-1}}$.
783: The parameters adopted for the hot wind component in V1016~Cyg are
784: $v_c= -10{\rm\ km\ s^{-1}}$ and $\Delta v_c = 9{\rm\ km\ s^{-1}}$
785: and $v_c= -7{\rm\ km\ s^{-1}}$ and $\Delta v_c = 6{\rm\ km\ s^{-1}}$
786: for HM~Sge.
787:
788: \begin{figure}
789: %\epsfig{file=fit_acc.ps, scale = 0.67, angle = 270}
790: %\epsfig{file=f7.eps, scale = 0.67, angle = 270}
791: \plotone{f7.eps}
792: \caption{Line profiles from our Monte Carlo
793: calculations for various velocities $v_0$ at the disk outer rim.
794: Dots show the BOES spectra and lines represent our Monte
795: Carlo results. Solid lines show our best fit profiles shown in
796: Fig.~6.
797: }
798: \label{fitacc}
799: \end{figure}
800:
801: In Fig.~\ref{fitacc}
802: we present the line profiles for various velocities $v_o$ at the disk outer
803: edge ranging from $24{\rm\ km\ s^{-1}}$ to $33{\rm\ km\ s^{-1}}$
804: with $v_\infty, v_c$ and $\Delta v_c$ fixed to the best fit values
805: of Fig.~(\ref{bestfit}).
806: The middle panel shows the best fit line profiles in Fig.~\ref{bestfit},
807: whereas the top
808: and bottom panels show line profiles with faster and slower disks,
809: respectively. Considering the much poorer fits shown in the top and
810: bottom panels, we may conclude that the outer rim velocities are
811: constrained to be in the range $v_o=30\pm3{\rm\ km\ s^{-1}}$
812: for V1016~Cyg and $v_o=26\pm 2{\rm\ km\ s^{-1}}$ for HM~Sge.
813: Adopting a typical white dwarf mass of $M_{WD}=0.7{\rm\ M_\odot}$,
814: the main O~VI emission region resides from the hot white dwarf component
815: at $\sim 0.7$ AU for V1016~Cyg and $\sim 0.9$ AU for HM~Sge.
816:
817:
818: Fig.~\ref{fitwind} shows the Monte Carlo line profiles
819: computed for various terminal giant wind velocities $v_\infty$
820: from $8{\rm\ km\ s^{-1}}$ to $13{\rm\ km\ s^{-1}}$
821: also with $v_o, v_c$ and $\Delta v_c$ fixed to the best fit
822: values introduced in Fig.~(\ref{bestfit}).
823: Similarly as in Fig.~\ref{fitacc}, the middle panels show our best
824: fit profiles in Fig.~\ref{bestfit} for comparison.
825: As is discussed in the previous section, larger giant wind velocities
826: tend to erase the double peak structure, whereas overall sharper
827: profiles result from smaller wind velocities yielding poor fits at the
828: outer parts of the observed profiles.
829: It is also interesting to note that the wind terminal velocity $v_\infty \sim
830: 10{\rm\ km\ s^{-1}}$ for both V1016~Cyg and HM~Sge
831: is quite similar to the escape velocity of a typical giant star.
832:
833: \begin{figure}
834: %\epsfig{file=fit_wind.ps, scale = 0.67, angle = 270}
835: %\epsfig{file=f8.eps, scale = 0.67, angle = 270}
836: \plotone{f8.eps}
837: \caption{Line profiles from our Monte Carlo
838: calculations for various terminal giant wind velocities $v_\infty$
839: with $v_o, v_c$ and $\Delta v_c$ fixed to the best fit values of
840: Fig.~(\ref{bestfit}). Dots show the BOES spectra
841: and lines represent our Monte Carlo results. The best fit profiles
842: in Fig. 6 are shown by solid lines.
843: Larger values of $v_\infty$ tend to yield better fits at far wing parts
844: but erase the double peak structure in the main parts.
845: }
846: \label{fitwind}
847: \end{figure}
848:
849: \subsection{Hot Spots}
850:
851: \begin{figure}
852: %\epsfig{file=spot.ps, scale = 0.67, angle = 270}
853: %\epsfig{file=f9.eps, scale = 0.67, angle = 270}
854: \plotone{f9.eps}
855: \caption{Monte Carlo line profiles from a Keplerian disk with a hot spot.
856: The solid lines show our Monte Carlo data and the dots show the BOES
857: data. The dotted lines show the best fit profiles given
858: in Fig.~\ref{bestfit}. The fitting parameters
859: are $v_o=33{\rm\ km\ s^{-1}}$ and $v_\infty = 9{\rm\ km\ s^{-1}}$
860: for V1016~Cyg and $v_o = 27{\rm\ km\ s^{-1}}$ and $v_\infty=
861: 9{\rm\ km\ s^{-1}}$ for HM~Sge.
862: }
863: \label{hotspot}
864: \end{figure}
865:
866: As is well known from spectroscopic studies
867: of cataclysmic variables, an accretion disk may have a bright spot
868: which is formed by the impact of the accretion stream near the outer
869: edge of the accretion disk (e.g. Warner 1995).
870: The existence of a similar feature in
871: symbiotic systems is not clear. In this subsection, we investigate
872: the effect of the introduction of a hot spot in a Keplerian disk
873: on the line profile.
874:
875: In our Monte Carlo code, photons generated in the
876: specified region satisfying $r>0.8r_o$ and $|\phi+0.7\pi|<0.1\pi$ is assigned
877: a statistical weight of 4.8, so that this region is 4.8 times brighter
878: than other part of the emission region. This mimics the bright spots well
879: known in cataclysmic variables, for which the
880: accretion process takes place via Roche lobe overflow.
881: Here, we do not consider the hot wind component that provided central
882: absorption part in Fig.~\ref{bestfit}.
883:
884: Fig.~\ref{hotspot} shows our profiles. The solid lines show the profiles
885: from a Keplerian disk with a hot spot, whereas the dots show the BOES
886: data. For the sake of comparisons, we also show the best fit profiles given
887: in Fig.~\ref{bestfit} by dotted lines. The fitting parameters for
888: V1016~Cyg are $v_o=33{\rm\ km\ s^{-1}}$ and $v_\infty = 9{\rm\ km\ s^{-1}}$.
889: For HM~Sge, they are $v_o = 27{\rm\ km\ s^{-1}}$ and $v_\infty=
890: 9{\rm\ km\ s^{-1}}$. It should be noted that the disk outer rims move
891: slightly faster in this model than the ones considered
892: in the case of the best fit model.
893:
894:
895: The profiles provide good fits to the outer parts
896: $\lambda<6824{\rm\ \AA}$ and $\lambda>6830{\rm\ \AA}$, whereas poor
897: fits are obtained near center. The introduction of a hot spot enhances
898: the red peak. However, the conspicuous central dips in the observed data
899: are not obtained in pure Keplerian disk models with $v_o >25{\rm\ km\ s^{-1}}$
900: for $v_\infty >8{\rm\ km\ s^{-1}}$. Lowering $v_\infty$ makes the fit worse
901: by narrowing the widths of the two peaks.
902:
903: \section{Summary and Discussion}
904:
905: In this work, we present our profile analysis of Raman scattered
906: O~VI 6825 in the symbiotic stars V1016~Cyg and HM~Sge, assuming that
907: O~VI line photons are generated from a thin Keplerian disk with
908: a scattering region between the white dwarf and the Mira giant.
909: The neutral scattering region is assumed to be a spherical
910: slow stellar wind with terminal wind velocity $v_\infty$
911: with an ionization front approximated by a hyperboloid.
912: The best fitting parameters are $v_o=30{\rm\ km\ s^{-1}}$, $v_\infty=
913: 11{\rm\ km\ s^{-1}}$ and $v_w = -10{\rm\ km\ s^{-1}}$ for V1016~Cyg
914: and $v_o=26{\rm\ km\ s^{-1}}$, $v_\infty = 10{\rm\ km\ s^{-1}}$
915: and $v_w=-7{\rm\ km\ s^{-1}}$ for HM~Sge.
916:
917: The overall fitting quality is quite insensitive to our choice of
918: the inner radius of the O~VI emission region and the functional
919: dependence of emissivity on radius. Furthermore, the profile shape
920: is little affected by the exact shape
921: of the ionization structure, which is determined by the mass loss rate
922: of the Mira component and the ionizing luminosity from the white dwarf
923: component. However, the ionization structure mainly affects the total
924: flux of Raman scattered 6825. In this point, an independent study
925: should be added to current work in order to adequately probe
926: the ionization structure and mass loss and transfer processes in
927: symbiotic stars (e.g. Jung \& Lee 2004).
928:
929: Although the introduction of a hot spot in our Keplerian disk
930: model failed to improve the profile fitting quality in the case
931: of V1016~Cyg and HM~Sge,
932: this should not mean to exclude the possibility of the asymmetrical
933: emission in symbiotic stars. As Harries \& Howarth (1996) reported,
934: triple peak structures in Raman scattered 6825 in symbiotics are
935: more prevalent than double peak profiles in symbiotic stars.
936: It is quite probable that more complicated profiles may be
937: fitted using more ingredients in the emission and/or scattering regions.
938:
939: The resonance scattering O~VI component considered in section 4.3
940: may be envisioned as a part of ionized giant
941: stellar wind, in the vicinity of the inner Lagrange point.
942: In order to scatter a significant fraction of slightly blue O~VI photons, the
943: region should move toward the main O~VI line emission region
944: around the white dwarf. We may note that the bestfit speed $v_c$ in
945: Fig.~\ref{bestfit} is slightly less than the giant wind terminal
946: velocity. Because $v_c$ is the velocity component along the direction
947: connecting the giant and the white dwarf, we may interpret that the
948: resonantly scattering O~VI region passes through the inner
949: Lagrange point making an angle
950: \begin{equation}
951: \theta_i \simeq \sin^{-1}{0.8} = 50^\circ
952: \end{equation}
953: under the assumption that the speed of the region is coincident with
954: the wind terminal velocity $v_\infty$.
955: With the wide binary orbit of $50-80$ AU the orbital speed of
956: the giant component in both V1016~Cyg and HM~Sge is comparable
957: to $v_c$. Although we only consider the spherical wind around
958: the Mira type giant component in the current work, it is also a
959: possibility that the wind around the giant component
960: also possesses an azimuthal component, in which case this inference
961: of $\theta_i$ should be corrected appropriately. Refined hydrodynamical
962: studies may shed more light on the mass transfer processes in symbiotic
963: stars.
964:
965: The red wing excess in V1016~Cyg that is not well fitted by our model
966: may indicate the existence of another emission component and/or
967: a neutral scattering region. Spectropolarimetric studies by
968: Harries \& Howarth (1996) show that the far red wing part is polarized
969: in the direction
970: perpendicular to that for the remainder part. These facts are consistent
971: when the additional emission and/or neutral component is receding
972: with a velocity of several
973: tens of ${\rm km\ s^{-1}}$ perpendicular to the binary orbital plane.
974:
975: It is unclear that this component is associated with the bipolar
976: structure of these objects. Solf (1983, 1984) detected a bipolar outflow
977: in [N II] lines in HM~Sge.
978: The velocity scale of several tens of km s$^-1$
979: is much smaller than the usual velocity scale associated with the fast
980: hot wind that may emanate from a deep inner region of an accretion disk
981: in the vicinity of the hot star. Schmid et al. (1999) reported
982: the existence of broad wing components in O~VI 1032, 1038 observed with
983: ORFEUS. Even though the broad wings are consistent with the electron
984: scattering origin, a fast bipolar wind around the hot component
985: can also be a candidate. It is an interesting possibility that
986: the component responsible for resonance scattering of O~VI is a clumpy
987: one driven by the fast hot wind moving in the polar directions.
988: More interesting results are expected when
989: photoionization calculations and hydrodynamical studies can be combined
990: with spectropolarimetric polarimetric observations of these symbiotic objects.
991:
992:
993: %He II Raman Scattered Lines
994:
995: It is also interesting that these two symbiotic stars exhibit He II Raman
996: scattered features blueward of H Balmer lines. In particular, Raman scattered
997: He II 4850 features of V1016 Cyg and HM Sge exhibit no multiple peak
998: structures (e.g. Jung \& Lee 2004, Birriel 2004).
999: Because the cross section for He II Raman scattering is larger than
1000: that for O~VI Raman scattering by two orders of magnitude, we may expect
1001: that the scattering region is much more extended than the Raman O~VI
1002: counterpart.
1003:
1004:
1005: The profiles of Raman O~VI 6825 in V1016~Cyg and
1006: HM~Sge are totally different from other high ionization lines including
1007: He~II. For example, He~II~4686 emission lines of V1016~Cyg and HM~Sge
1008: are single-peaked with slight blue asymmetry.
1009: Robinson et al. (1994) considered line profiles from accretion disks
1010: in symbiotic stars, and it appears to be difficult to infer the existence
1011: of an accretion disk in symbiotic stars from these emission lines.
1012: It may be that the He~II emission region is more extended than the O~VI
1013: emission region. It is beyond the scope of this work to model other
1014: emission lines and more sophisticated models including photoionization
1015: and hydrodynamics are required.
1016:
1017: The disparity of the profiles in Raman scattered O~VI 6825 and Raman
1018: scattered O~VI 7088 can be an important clue to the structure of the
1019: accretion disk in symbiotic stars. A number of researchers
1020: pointed out the relative weakness in the blue part of Raman O~VI 7088
1021: (e.g. Schmid et al. 1999, Harries \& Howarth 1996). The profile
1022: difference in resonance doublets of $S_{1/2}-P_{1/2,3/2}$ was reported
1023: in symbiotic stars and young planetary nebulae (e.g. Feibelman 1983).
1024:
1025:
1026: \acknowledgements
1027: We thank the staff at the Bohyunsan Observatory with our particular
1028: gratitude to Kang Min Kim and Byung Cheol Lee. We are also grateful to
1029: Hwankyung Sung for useful discussions on our spectroscopic
1030: observation. We thank an anonymous referee, whose suggestions improved
1031: the presentation of the current paper. This work is a
1032: result of research activities of the Astrophysical Research
1033: Center for the Structure and Evolution of the Cosmos (ARCSEC)
1034: funded by the Korea Science and Engineering Foundation.
1035:
1036: \appendix
1037:
1038: \section{The Ionization Fronts}
1039:
1040: We describe in detail the photoionization calculation for the ionization
1041: fronts, which was briefly discussed in section 2.
1042:
1043: The distance from the giant is measured in units of $R_*$, the radius of the
1044: giant, by $\rho = r/R_*$. Let $\rho_i$ be the separation of the giant
1045: and the white dwarf. Along a path from the white dwarf making an angle $\phi$
1046: with the line connecting the giant, the physical distance is measured
1047: by the parameter $s$ satisfying
1048: \begin{equation}
1049: s =
1050: \cases{\rho_i\cos\phi - \sqrt{\rho^2 -b^2} \quad {\rm for}\
1051: s\le \rho_i\cos\phi \cr
1052: \rho_i\cos\phi + \sqrt{\rho^2 -b^2} \quad {\rm for}\
1053: s\ge \rho_i\cos\phi \cr
1054: }
1055: \end{equation}
1056: where $b=\rho_i\sin\phi$ is the impact parameter from the giant to the path of
1057: the ionization radiation from the white dwarf.
1058:
1059: With the density law
1060: \begin{equation}
1061: n(\rho) = {n_0\over \rho(\rho-1)},
1062: \end{equation}
1063: where the distance $r$ from the giant center is measured in units of
1064: $R_*$ by the parameter $\rho=r/R_*$.
1065:
1066: The recombination rate along the path of ionization radiation
1067: originating from the white dwarf is proportional to the density squared
1068: along the path, given by
1069: \begin{equation}
1070: I = \int n(\rho)^2 s^2 ds.
1071: \label{int1}
1072: \end{equation}
1073: In terms of $\rho$, this integral can be re-written as
1074: \begin{eqnarray}
1075: I &=& \int{(\rho_i\cos\phi\pm\sqrt{\rho^2-b^2})^2\over \rho(\rho-1)^2
1076: \sqrt{\rho^2-b^2}} d\rho
1077: \nonumber \\
1078: &=&\rho_i^2\cos^2\phi I_1+I_2\pm 2\rho_i I_3,
1079: \end{eqnarray}
1080: We define the three integrals as
1081: \begin{eqnarray}
1082: I_1 &=& \int{d\rho\over \rho(\rho-1)^2\sqrt{\rho^2-b^2}}
1083: \nonumber \\
1084: I_2 &=& \int{\sqrt{\rho^2-b^2}d\rho\over \rho(\rho-1)^2}
1085: \nonumber \\
1086: I_3 &=& \int{d\rho\over \rho(\rho-1)^2} .
1087: \end{eqnarray}
1088: The integral $I_3$ is elementary and is given by
1089: \begin{equation}
1090: I_3=\ln{\rho\over \rho-1}-{1\over \rho-1}.
1091: \end{equation}
1092:
1093: Considering the substitution
1094: \begin{equation}
1095: \rho = b\cosh\theta
1096: \label{rhotheta}
1097: \end{equation}
1098: the integrals under consideration become
1099: \begin{eqnarray}
1100: I_1 &=& \int d\theta
1101: \left[
1102: {1\over b\cosh\theta}-{1\over b\cosh\theta-1}
1103: +{1\over (b\cosh\theta-1)^2}
1104: \right]
1105: \nonumber \\
1106: I_2&=& \int d\theta b^2(\cosh^2\theta-1)
1107: \left[
1108: {1\over b\cosh\theta}-{1\over b\cosh\theta-1}
1109: +{1\over (b\cosh\theta-1)^2}
1110: \right]
1111: \label{intj}
1112: \end{eqnarray}
1113:
1114: In order to obtain the closed form expressions
1115: of the integrals $I_1$ and $I_2$, we define the two integrals
1116: $K$ and $L$ as
1117: \begin{eqnarray}
1118: K &=& \int{d\theta \over \cosh\theta-\epsilon}
1119: \nonumber \\
1120: L &=& \int{d\theta \over (\cosh\theta-\epsilon)^2} ,
1121: \end{eqnarray}
1122: with $\epsilon=1/b$.
1123:
1124: For $\epsilon<1$, making use of the half angle formula
1125: for $\cosh\theta$ and the substitution
1126: \begin{equation}
1127: \tanh{\theta\over 2}=\sqrt{1-\epsilon\over 1+\epsilon} u,
1128: \end{equation}
1129: we obtain
1130: \begin{equation}
1131: K = {2\over\sqrt{1-\epsilon^2}}
1132: \tan^{-1}\sqrt{1+\epsilon\over 1-\epsilon} \tanh{\theta\over2}.
1133: \end{equation}
1134:
1135: On the other hand, for $\epsilon>1$, then the substitution we consider is
1136: \begin{equation}
1137: \tanh{\theta\over 2}=\sqrt{\epsilon-1\over \epsilon+1} v,
1138: \end{equation}
1139: which leads us to
1140: \begin{equation}
1141: K = {2\over\sqrt{\epsilon^2-1}}\int {dv\over v^2-1} =
1142: -{2\over\sqrt{\epsilon^2-1}}\coth^{-1}\sqrt{\epsilon+1\over \epsilon-1}
1143: \tanh{\theta\over2}.
1144: \end{equation}
1145:
1146: We also consider the following relation
1147: \begin{equation}
1148: {d\over d\theta}\left({\sinh\theta\over \cosh\theta-\epsilon}\right)
1149: =-{\epsilon\over\cosh\theta-\epsilon}+{1-\epsilon^2
1150: \over(\cosh\theta-\epsilon)^2},
1151: \end{equation}
1152: from which we obtain
1153: \begin{equation}
1154: L = {\epsilon\over 1-\epsilon^2}K+{1\over1-\epsilon^2}{\sinh\theta
1155: \over \cosh\theta-\epsilon}.
1156: \end{equation}
1157:
1158: The integrals $I_1, I_2$ in Eq.~(\ref{intj}) can be finally expressed as
1159: \begin{eqnarray}
1160: I_1 &=& {1\over b}\tan^{-1}\sinh\theta-{K\over b}+{L\over b^2}
1161: \nonumber \\
1162: I_2&=&-b\tan^{-1}\sinh\theta +{(b^2+1)\over b}K-{(b^2-1)\over b^2}L.
1163: \end{eqnarray}
1164:
1165: Combining these results, for $\epsilon<1$, an explicit expression
1166: of the integral $I$ is given by
1167: \begin{eqnarray}
1168: I &=& \rho_i\cos2\phi\csc\phi\tan^{-1}\sinh\theta
1169: +{\epsilon^2-\csc^2\phi \over 1-\epsilon^2}{\sinh\theta
1170: \over \cosh\theta - \epsilon}
1171: \nonumber \\
1172: &+&
1173: {4\epsilon^2\cot^2\phi-2\epsilon^2+2\csc^2\phi \over
1174: \epsilon(1-\epsilon^2)^{3/2}}
1175: \tan^{-1}\sqrt{1+\epsilon\over 1-\epsilon}
1176: \tanh{\theta\over2}
1177: \nonumber \\
1178: & \pm&
1179: 2\rho_i\cos\phi
1180: \left(\ln{\rho\over\rho-1}
1181: -{1\over\rho-1}\right).
1182: \label{strom1}
1183: \end{eqnarray}
1184:
1185: Also, for $\epsilon>1$, we have
1186: \begin{eqnarray}
1187: I &=& \rho_i\cos2\phi\csc\phi\tan^{-1}\sinh\theta
1188: +{\epsilon^2-\csc^2\phi \over 1-\epsilon^2}{\sinh\theta
1189: \over \cosh\theta - \epsilon}
1190: \nonumber \\
1191: &-&
1192: {4\epsilon^2\cot^2\phi-2\epsilon^2+2\csc^2\phi \over
1193: \epsilon(\epsilon^2-1)^{3/2}}
1194: \coth^{-1}\sqrt{\epsilon+1\over \epsilon-1}
1195: \tanh{\theta\over2}
1196: \nonumber \\
1197: &-&
1198: 2\rho_i\cos\phi
1199: \left(\ln{\rho\over\rho-1}
1200: -{1\over\rho-1}\right).
1201: \end{eqnarray}
1202:
1203: Noting Eq.(\ref{rhotheta}), we may immediately see that Eq.(\ref{strom1})
1204: reduces to Eq.(\ref{eqifront1}).
1205: In Fig.~\ref{ifront}, we show ionization fronts
1206: for a range of ionizing luminosities characterized by the parameter
1207: $X$. The cool giant is located at the origin and the white dwarf
1208: is at $(\rho_i=80,0)$. We also approximate the
1209: ionization corresponding to $X=0.02$ by a hyperbola
1210: given by $y^2=6.5(x-40)^2-400$.
1211:
1212:
1213: \begin{thebibliography}{}
1214: \bibitem[Allen (1980)]{all80} Allen, D. A., 1980, \mnras, 190, 75
1215:
1216: \bibitem[Birriel (2004)]{bir04} Birriel, J. J., 2004, \apj, 612, 1136
1217:
1218: \bibitem[Birriel, Espey, \& Schulte-Ladbeck (2000)]{bir00} Birriel,
1219: J. J., Espey, B. R., \& Schulte-Ladbeck, R. E., 2000, \apj, 545, 1020
1220:
1221: \bibitem[Brocksopp et al. (2002)]{bro02} Brocksopp, C., Bode, M. F.,
1222: Eyres, S. P. S., Crocker, M. M., Davis, R. \& Taylor, A. R., 2002,
1223: \apj, 571, 947
1224:
1225: \bibitem[Brocksopp et al. (2004)]{bro04} Brocksopp, C., Sokoloski,
1226: J. L., Kaiser, C., Richards, A. M., Muxlow, T. W. B., \& Seymour,
1227: N., 2004, \mnras, 347, 430
1228:
1229: %\bibitem[Corradi \& Schwarz (1995)]{cor95} Corradi, R. L. M., \&
1230: % Schwarz, H. E., 1995, \aap, 293, 871
1231:
1232: \bibitem[Dokuchaeva 1976]{dok76} Dokuchaeva, O. D., 1976, IBVS, 1189
1233:
1234: \bibitem[Eyres et al. 2001]{eyr01} Eyres, S. P. S., Bode, M. F.,
1235: Taylor, A. R., Crocker, M. M., Davis, R. J., 2001, \apj, 551,
1236: 512
1237:
1238: \bibitem[Feibelman (1983)]{fei83} Feibelman, W. A., 1983, \aap, 122,
1239: 335
1240:
1241: \bibitem[Fitzgerald et al. 1966]{fit66} Fitzgerald, M. P., Houk, N.,
1242: McCuskey, S. W., \& Hoffleit, D., 1966, \apj, 144, 1135
1243:
1244: \bibitem[Groves et al. 2002]{gro02} Groves, B., Dopita, M. A.,
1245: Williams, R. E., Hua, C.-T., 2002, PASA, 19, 425
1246:
1247: \bibitem[Harries \& Howarth 1996]{har96} Harries, T. J. \& Howarth, I. D.
1248: 1996, \aaps, 119, 61
1249:
1250: \bibitem[Harries \& Howarth 1997]{har97} Harries, T. J. \& Howarth, I. D.
1251: 1997, \aaps, 121, 15
1252:
1253: \bibitem[Iben \& Tutukov 1996]{ibe96} Iben, I. \& Tutukov, A. V.,1996,
1254: \apjs, 105, 145
1255:
1256: \bibitem[Jung \& Lee 2004]{jun04} Jung, Y. -C. \& Lee, H. -W., 2004,
1257: \mnras, 355, 221
1258:
1259: \bibitem[Kenyon 1986]{ken86} Kenyon, S., 1986,
1260: The Symbiotic Stars (Cambridge University Press: New York)
1261:
1262: \bibitem[Kim et al. 2002]{kim02} Kim, K. -M., Jang, B. -H., Han, I.,
1263: Jang, J. G., Sung, H. C., Chun, M. -Y., Hyung, S., Yoon, T. -S., Vogt, S.,
1264: 2002, Journal of the Korean Astronomical Society, 35, 221
1265:
1266: \bibitem[Lee et al. 2006a]{lee06a} Lee, B. -C., Galazutdinov, G. A.,
1267: Han, I., Kim, K. -M., Yushchenko, A. V., Kim, J., Tsymbal, V., Park, M. -G.,
1268: 2006, PASP, 118, 636 (Lee et al. 2006a)
1269:
1270: \bibitem[Lee \& Lee (1997a)]{lee97a} Lee, H. -W. \& Lee, K. W.
1271: 1997, \mnras, 287, 211
1272:
1273: \bibitem[Lee \& Lee (1997b)]{lee97b} Lee, K. W. \& Lee, H. -W.
1274: 1997, \mnras, 292, 573
1275:
1276: \bibitem[Lee \& Park (1999)]{lee99} Lee, H. -W. \& Park, M. -G.,
1277: 1999, \apj, 515, L89
1278:
1279: \bibitem[Lee et al. (2003)]{lee03} Lee, H. -W., Sohn, Y. -J., Kang,
1280: Y. W., \& Kim, H. -I., 2003, \apj, 598, 553
1281:
1282: \bibitem[Lee et al. (2006)]{lee06b} Lee, H. -W., Jung, Y. -C., Song, I. -O.,
1283: Ahn, S. -H., 2006, \apj, 636, 1045 (Lee et al. 2006b)
1284:
1285: %\bibitem[Livio \& Warner 1984]{liv84} Livio, M., \& Warner, B., 1984,
1286: % The Observatory, 104, 152
1287:
1288: \bibitem[Luthardt 1992]{lut92} Luthardt, R., 1992, Reviews of Modern Astronomy,
1289: 5, 38
1290:
1291: \bibitem[Mastrodemos \& Morris (1998)]{mas98} Mastrodemos, N., \& Morris,
1292: M., 1998, \apj, 497, 303
1293:
1294: \bibitem[M\"urset et al. 1991]{mur91} M\"urset, U., Nussbaumer, H.,
1295: Schmid, H. M., Vogel, M., 1991, \aap, 248, 458
1296:
1297: \bibitem[Nussbaumer, Schmid \& Vogel (1989)]{nus89} Nussbaumer, H.,
1298: Schmid, H. M., \& Vogel, 1989, \aap, 211, L27
1299:
1300: \bibitem[P\'equignot et al. (1997)]{peq97} P\'equignot, D., Baluteau,
1301: J. -P., Morisset, C., Boisson, C., 1997, \aap, 323, 217
1302:
1303: %\bibitem[Raymond et al. (2003)]{ray03} Raymond, J. C., Ghavamian, P.,
1304: % Sankrit, R., Blair, W. P., Curiel, S., 2003, \apj, 584, 770
1305:
1306: \bibitem[Robinson et al. (1994)]{rob94} Robinson, K., Bode, M. F.,
1307: Skopal, A., Ivison, R. J., \& Meaburn, J., 1994, \mnras, 269, 1
1308:
1309: %\bibitem[Sadeghpour \& Dalgarno (1992)]{sad92} Sadeghpour, H. R., \&
1310: % Dalgarno, A., 1992, J. Phys. B: At. Mol. Opt. Phys., 25, 4801
1311:
1312: %\bibitem[Sankrit \& Blair (2002)]{san02} Sankrit, R., \& Blair, W. P.,
1313: % 2002, \apj, 565, 297
1314:
1315: \bibitem[Schild \& Schmid (1996)]{schl96} Schild, H. \& Schmid, H. M.,
1316: 1996, \aap, 310, 211
1317:
1318: \bibitem[Schmid (1989)]{sch89} Schmid, H. M., 1989, \aap,
1319: 211, L31
1320:
1321: \bibitem[Schmid (1996)]{sch96} Schmid, H. M., 1996, \mnras,
1322: 282, 511
1323:
1324: \bibitem[Schmid (1998)]{sch98} Schmid, H. M., 1998, Reviews of Modern
1325: Astronomy, 11, 297
1326:
1327: \bibitem[Schmid et al. (1999)]{sch99} Schmid, H. M. et al., 1999, \aap,
1328: 348, 950
1329:
1330: \bibitem[Schmid et al. (2000)]{sch00} Schmid, H. M. Corradi, R., Krautter, J.
1331: and Schild, H., 2000, \aap, 355, 261
1332:
1333: \bibitem[Schmid (2001)]{sch01} Schmid, H. M., 2001, ASP Conference
1334: Series, 242, 347, 'Eta Carinae \& Other Mysterious Stars'
1335: ed T. Gull, S. Johansson \& K. Davidson
1336:
1337: %\bibitem[Schwarz \& Monteiro (2003)]{sch03} Schwarz, H. E., \&
1338: % Monteiro, H., 2003, Rev. Mexicana Astron. Astrofis., 15, 23
1339:
1340: \bibitem[Sokoloski, Bildsten \& Ho 2001]{sok01} Sokoloski, J. L., Bildsten, L.,
1341: Ho, W. C. G., 2001, MNRAS, 326, 553
1342:
1343: \bibitem[Solf (1983)]{sol83} Solf, J., 1983, \apj, 266, L113
1344:
1345: \bibitem[Solf (1984)]{sol84} Solf, J., 1984, \aap, 139, 296
1346:
1347: \bibitem[Taylor \& Seaquist (1984)]{tay84} Taylor, A. R., \& Seaquist,
1348: E. R., 1984, \apj, 286, 263
1349:
1350: \bibitem[Theuns \& Jorrissen (1993)]{the93} Theuns, T. \& Jorrissen, A.,
1351: 1993, MNRAS, 265, 946
1352:
1353: \bibitem[van Groningen (1993)]{van93} Van Groningen, E., 1993, \mnras,
1354: 264, 975
1355:
1356: %\bibitem[Wheatley, Mukai \& de Martino (2003)]{whe03} Wheatley, P. J.,
1357: % Mukai, K., \& de Martino, D., 2003, \mnras, 346, 855
1358:
1359: \bibitem[Warner 1995]{war95} Warner, B., 1995, Cataclysmic Variable Stars,
1360: Cambridge University Press, Cambridge
1361:
1362: \bibitem[Whitelock 1987]{whi87} Whitelock, P. A., 1987, PASP, 99, 573
1363:
1364: \bibitem[Zhang et al. 2005]{zha05} Zhang, Y., Liu, X.-W., Luo, S.-G.,
1365: P\'equignot, D., Barlow, M. J., 2005, \aap, 442, 249
1366:
1367: \end{thebibliography}
1368:
1369:
1370: \end{document}
1371: