1: %\documentclass{aastex}
2:
3: % \documentclass[preprint]{aastex}
4: \documentclass[preprint2]{aastex}
5:
6: \newcommand{\vdag}{(v)^\dagger}
7: \newcommand{\myemail}{kawabtkj@hiroshima-u.ac.jp}
8:
9:
10: \shorttitle{SPECTROPOLARIMETRY OF R CrB IN 1998--2003}
11: \shortauthors{Kawabata et al.}
12:
13: \begin{document}
14:
15: \title{SPECTROPOLARIMETRY OF R CORONAE BOREALIS IN 1998--2003:
16: DISCOVERY OF TRANSIENT POLARIZATION AT MAXIMUM BRIGHTNESS}
17:
18: \author{K.~S.~Kawabata\altaffilmark{1,2},
19: Y.~Ikeda\altaffilmark{3,2},
20: H.~Akitaya\altaffilmark{4,2},
21: M.~Isogai\altaffilmark{1,2},
22: K.~Matsuda\altaffilmark{5},
23: M.~Matsumura\altaffilmark{6},
24: O.~Nagae\altaffilmark{7}, and
25: M.~Seki\altaffilmark{8}
26: }
27:
28: \altaffiltext{1}{Hiroshima Astrophysical Science Center,
29: Hiroshima University, 1-3-1 Kagamiyama, Higashi-Hiroshima,
30: Hiroshima 739-8526, Japan;
31: kawabtkj@hiroshima-u.ac.jp, isogai@hiroshima-u.ac.jp}
32:
33: \altaffiltext{2}{Visiting Astronomer, Okayama Astrophysical
34: Observatory of National Astronomical Observatory of Japan (NAOJ)}
35:
36: \altaffiltext{3}{Photocoding, Higashi-Hashimoto 3-16-8-101, Sagamihara,
37: Kanagawa 229-1104, Japan;
38: ikeda@photocoding.com}
39:
40: \altaffiltext{4}{Division of Optical and Infrared Astronomy,
41: NAOJ, Osawa 2-21-1, Mitaka, Tokyo 181-8588, Japan;
42: akitaya@optik.nao.ac.jp}
43:
44: \altaffiltext{5}{Nishi-Harima Astronomical Observatory, Sayo-cho,
45: Hyogo 679-5313, Japan;
46: matsuda@nhao.go.jp}
47:
48: \altaffiltext{6}{Faculty of Education, Kagawa University,
49: Saiwai-cho 1-1, Takamatsu 760-8522, Japan;
50: matsu@ed.kagawa-u.ac.jp}
51:
52: \altaffiltext{7}{Department of Physical Science, School of Science,
53: Hiroshima University, 1-3-1 Kagamiyama, Higashi-Hiroshima,
54: Hiroshima 739-8526, Japan;
55: nagae@hep01.hepl.hiroshima-u.ac.jp}
56:
57: \altaffiltext{8}{Astronomical Institute, Graduate School of Science,
58: Tohoku University, Aramaki, Aoba-ku, Sendai 980-8578, Japan;
59: seki@astr.tohoku.ac.jp}
60:
61:
62: \begin{abstract}
63: We present an extended optical spectropolarimetry of R CrB
64: from 1998 January to 2003 September.
65: The polarization was almost constant in the phase of maximum
66: brightness, being consistent with past observations.
67: We detected, however, temporal changes of polarization ($\sim 0.5$ \%)
68: in 2001 March and August, which were the first detection of
69: large polarization variability in R CrB near maximum brightness.
70: The amplitude and the position angle of the `transient polarization'
71: were almost constant with wavelength in both two events.
72: There was a difference by about 20 degrees in the position angle
73: between the two events.
74: Each event could be explained by light scattering due to
75: short-lived dust puff occasionally ejected off the line of sight.
76: The flatness of the polarization against the wavelength
77: suggests that the scatterer is a mixture of dust grains
78: having various sizes.
79: The rapid growth and fading of the transient polarization
80: favors the phenomenological model of dust formation near the
81: stellar photosphere (e.g., within two stellar radii)
82: proposed for the time evolution of brightness
83: and chromospheric emission lines during deeply
84: declining periods, although the fading timescale can hardly
85: be explained by a simple dispersal of expanding dust puff
86: with a velocity of $\sim 200-350$ km s $^{-1}$.
87: Higher expansion velocity or some mechanism to destroy the
88: dust grains should be needed.
89: \end{abstract}
90:
91: \keywords{circumstellar matter --- dust, extinction ---
92: stars: individual (R Coronae Borealis) ---
93: stars: mass-loss --- techniques: polarimetric}
94:
95:
96: \section{INTRODUCTION}
97:
98: R Coronae Borealis (RCB) stars are hydrogen-deficient,
99: carbon-rich variables that undergo declines in visual
100: brightness up to 7 mag or more at irregular intervals.
101: The evolutionary pathways of these stars are still unclear,
102: with the white dwarf merger model and the final
103: helium-shell flash model being suggested for the
104: extreme abundances (e.g., \citealt{ibe96,sai02,cla07}).
105: The darkening phenomenon has been attributed to eclipse by clouds
106: of dust grains (dust puff) formed along the line of sight
107: \citep{lor34,oke39}.
108: Near-infrared excess, attributable to the dust emission
109: at an equivalent blackbody temperature of $\sim 900$ K,
110: was found for the prototype of this group, R CrB \citep{ste69},
111: and this excess is a general property of RCB stars \citep{fea73}.
112: The infrared radiation shows variations on timescales of
113: $\sim 3.5$ yr \citep{fea97b} and does not clearly correlate
114: with the optical brightness \citep{for72,fea97b,yud02}.
115: These indicate that a large amount of dust being produced
116: surrounds the star and that a smaller dust puff is formed
117: in a random direction per any one ejection event
118: (e.g., \citealt{ohn01,ohn03}).
119: Recent near-infrared adaptive optics observations with an
120: 8-m class telescope directly revealed the presence of multiple
121: dusty clouds in the vicinity of RY Sgr \citep{lav04}.
122:
123: The dust formation mechanism itself is, however, still one of the
124: fundamental problems in RCB stars (e.g., \citealt{cla96,fea97a,fea00}).
125: \citet{fad86} derived from a theoretical argument
126: a conclusion that the dust formation has to occur somewhere
127: $\sim 20 R_{\rm *}$ apart from the central star
128: where the local temperature becomes low enough to allow for
129: carbon grains to condense.
130: On the other hand, observational facts such as the time
131: evolution of chromospheric emission lines and timescales
132: of brightness recovering from declines are favorable to
133: other models in which dust forms in close proximity
134: ($\lesssim 2 R_{\rm *}$) to the central star \citep{cla92,whi93}.
135: To settle this discrepancy, some mechanism to promote
136: dust formation near RCB stars have been proposed.
137: \citet{woi96} suggested that the local gas temperature can be
138: substantially decreased by supercooling after the passage
139: of a shock wave driven by stellar pulsation.
140: \citet{asp96} proposed an episodic local density
141: inversions in the ionization layer of helium
142: due to a kind of radiative instability which
143: can promote a blob ejection.
144: \citet{fea97a} discussed the dust formation above the cool
145: regions of giant convection cells.
146: \citet{sok99} suggest that cool magnetic spots existing
147: inside of photosphere facilitate dust formation
148: above the spots after a shock wave passes.
149:
150: Polarimetry is a probe sensitive to circumstellar medium
151: near the stellar surface, and even to dust formation episodes
152: around RCB stars on very small spatial scales.
153: If dust forms actually near the photosphere
154: ($\lesssim 2 R_{\rm *}$), the angular size of the dust puff
155: should be more than $50\arcdeg$ for total eclipse of the stellar disk.
156: Such a wide dust puff is likely to scatter the light from the star
157: and may cause a net polarization more than $0.5$--$1$ \%,
158: when it locates along the direction nearly orthogonal to
159: the line of sight \citep{cod95}.
160: Therefore, it is natural to expect that RCB stars temporally show a
161: significant polarization at maximum brightness stage.
162: The wavelength dependence of the polarization would give
163: unique information about the ``newborn'' dust grain.
164: So far only small fluctuation of intrinsic polarization
165: less than 0.2 \% has been found for RCBs near maximum brightness
166: \citep{ser69,coy73,coy74,efi80,efi90,sta88,ros90,whi92a,tra94,
167: cla95,cla97,yud03,bie06},
168: except for \citet{efi80}'s observation on 1974 August 24
169: ($p_{\rm V} = 0.38\pm 0.03$ \%).
170:
171: In this paper, we present a new spectropolarimetry of R CrB
172: on 84 nights in 1998--2003.
173: We successfully found two events of temporal increase in
174: polarization by about 0.5 \%\ at maximum brightness stage.
175: We consider the period when the visual magnitude is within
176: 0.3 mag below the maximum $V$ magnitude ($V_{\rm max}$) as
177: ``maximum brightness'' phase and call the other period
178: ``decline'' phase.
179: We adopt $V_{\rm max} = 6.09$ which is the mean magnitude
180: during the apparent visual maximum of R CrB,
181: JD2,450,800--51,000 and 52,000--52,650 (Figure \ref{fig1}a).
182: The photometric data referred in this paper are obtained from
183: the VSNET\footnote{Variable Star NETwork;
184: \url{\tt http://www.kusastro.kyoto-u.ac.jp/vsnet/}}
185: and AAVSO\footnote{American Association of Variable Star
186: Observers; \url{\tt http://www.aavso.org/}} databases.
187:
188:
189: \section{OBSERVATIONS AND DATA REDUCTION}
190:
191: The data were obtained with the low-resolution spectropolarimeter,
192: HBS (an acronym of ``Henkou-Bunkou-Sokkou-Ki'' which
193: stands for spectropolarimeter in Japanese; \citealt{kaw99})
194: from 1998 January 28 to 2003 September 29 at the Dodaira Observatory
195: and the Okayama Astrophysical Observatory of NAOJ.
196: At the Dodaira Observatory, HBS was attached to the Cassegrain focus of
197: the 0.91 m telescope ($F/18$; $12\farcs 4$ mm$^{-1}$ at the focal plane).
198: At Okayama Astrophysical Observatory, HBS was attached to the
199: Cassegrain focus of either the 0.91 m telescope
200: ($F/13$; $17\farcs 2$ mm$^{-1}$) or the 1.88 m
201: telescope ($F/18$; $6\farcs 1$ mm$^{-1}$).
202: HBS has a superachromatic half-wave plate and a quarts Wollaston
203: prism; the orthogonally polarized spectra are simultaneously
204: recorded on either a front-illuminated type TI CCD
205: (1024$\times$1024 pixels, 12 $\micron$ square per pixel)
206: or a back-illuminated type SITe CCD (512$\times$512 pixels,
207: 24 $\micron$ square per pixel).
208: We used either the 1.4 mm$\phi$ circular hole (D1) or the
209: 0.2 mm $\times$ 1.4 mm rectangular hole (D2) as a focal
210: diaphragm.
211: Each diaphragm has two holes of the same dimension, and we put a
212: target star in one hole and the nearby sky in the other.
213: For the observations with D1 diaphragm, the spectral resolution
214: depends on the seeing size because the stellar image size is
215: much smaller than the diameter of the circular hole.
216: For the observations with D2, the spectral resolution also
217: depends on the stellar image size at the Okayama 0.91 m telescope,
218: while it does not at the other telescopes because the stellar
219: image size was usually larger than the narrower dimension of
220: the rectangular hole.
221: Nevertheless, the spectral resolution falls in the range between
222: 50\ \AA\ and 150\ \AA\ in the all observations.
223:
224: A unit of the observing sequence consisted of successive
225: integrations at four position angles (PA), $0\arcdeg$, $22\fdg 5$,
226: $45\arcdeg$, and $67\fdg 5$, of the half-wave plate.
227: The obtained images were processed using the reduction package
228: for HBS data, which was outlined in \citet{kaw99}.
229: Instrumental polarization was derived from unpolarized
230: standard star data obtained almost in each night.
231: The data were averaged in each observation run (typically
232: $\sim 10$ days for the Okayama 1.88 m telescope and 1-2 months for
233: the other telescopes). The level of instrumental
234: polarization $p_{\rm instr}$ was well expressed by a
235: smooth function of wavelength, and vectorially removed
236: from observed Stokes $Q$ and $U$ spectra.
237: %The $p_{\rm instr}$ showed different values between
238: %0.03 \% and 0.46 \% around 5500 \AA\; large variations
239: %were seen when we change the telescope and also when
240: %we replace of main optical train of HBS on 1999 August.
241: At any run, the stability ($1\sigma$) of $p_{\rm instr}$
242: was less than 0.05 \%.
243: The factor of instrumental depolarization was obtained from
244: observation through a Glan-Taylor prism.
245: This observation also gave us the wavelength-dependent
246: PA of the equivalent optical axis of the
247: superachromatic half-wave plate.
248: The zero point of the PA on the sky was determined
249: from observations of strongly polarized standard stars
250: listed in \citet{wol96}.
251: The instrumental polarization, the instrumental depolarization,
252: and the zero point of the PA were properly corrected for
253: in the reduction package.
254:
255: The observations are summarized in Table \ref{tbl-1}, in which
256: polarization data integrated over the synthetic
257: Johnson {\it V} band filter \citep{bes90} are shown.
258: For the observations with D2 diaphragm, we also show the
259: equivalent widths of the C$_{2} (0-0)$ 5165 \AA\ band in the table.
260: The flux was calibrated using spectrophotometry of
261: some standard stars \citep{tay84}.
262:
263:
264: \section{RESULTS}
265:
266: \subsection{Overview}
267:
268: Figure \ref{fig1} shows the time variation of polarization at
269: the synthetic {\it V} band, together with the visual light curve.
270: R CrB experienced three major photometric declines
271: in our observation period.
272: It demonstrates large polarization variability during declines;
273: for example, the polarization angle rotated by more than
274: 30$\arcdeg$ in 1999 December and the polarization amplitude
275: became over 5 \% in 2000 December.
276: On the other hand, R CrB near maximum brightness showed
277: nearly constant polarization (except for on 2001 March 9--12
278: and August 9).
279: These characteristics are consistent with past observations
280: (e.g., \citealt{efi80,sta88,cla95}).
281: A commonly-accepted explanation for the large polarization during
282: decline is that the unpolarized direct stellar light is heavily
283: obscured by the dust cloud (dust puff) along the line of sight
284: and the visible flux is predominantly scattered light.
285:
286: However, we detected temporal increase of
287: polarization two times even at maximum brightness stages;
288: the first event was observed
289: on 2001 March 9--12 and the second was on 2001 August 9.
290: In both times, the amplitude of polarization increased by
291: $\Delta p\simeq 0.5$ \% (Figure \ref{fig2}).
292: Although polarization usually increases accompanied with
293: the onset of major decline phases (Figure \ref{fig3}),
294: we cannot see any significant decline ($\Delta V\gtrsim 0.3$)
295: in the lightcurve around the two events.
296: Also, any remarkable change in flux spectra have not been
297: recognized as shown in Figure \ref{fig4}, and the equivalent width
298: of C$_{2}\ (0-0)$ 5165 \AA\ absorption band (Table \ref{tbl-1})
299: kept an almost constant value, 22--26 \AA , which is
300: consistent with the typical value of
301: R CrB near maximum brightness (17.5--43 \AA\ with an
302: average of 26.5 \AA ; \citealt{cla95}).
303: It should be noted, however, that there is possible
304: photometric variations accompanied with the polarimetric
305: variations.
306: The lightcurve shows possible shallow dips ($\sim$0.1 mag) a
307: few days prior to the polarization events (Figure \ref{fig2}),
308: which could be connected with small dust formation episodes.
309: We will discuss this temporal polarimetric activities
310: after subtraction of the constant polarization component.
311:
312:
313: \subsection{Constant Component of Polarization}
314:
315: It has been known that R CrB shows a nearly constant
316: polarization at maximum brightness.
317: \citet{sta88} averaged the polarization observed in a period
318: when no brightness fluctuations were seen, and derived a
319: constant component of $Q=-0.21$ \%, $U=-0.07$ \%
320: ($p=0.22$ \%, $\theta=99\arcdeg$) as the values
321: averaged over the optical wavelengths.
322: The same values were obtained from averaging subsequent
323: non-decline observations in 1990--1993 \citep{cla95}.
324: \citet{cla97} reported that the observation in 1995 March,
325: before the onset of the decline of 1995, showed a typical
326: Serkowski-type ISP function ($p_{\rm max}=0.24\pm 0.08$ \%,
327: $\lambda_{\rm max}=0.46\pm 0.01$ $\micron$, $\theta\sim 100\arcdeg$)
328: which was in good agreement with the previous estimations.
329: \citet{efi80} assumed the interstellar component of
330: $p_{\rm max}=0.2$ \% and $\theta = 110\arcdeg$ from polarization
331: of nearby field stars.
332: Similar technique also gives $p_{\rm max}=0.20$ \% and
333: $\theta = 97\arcdeg$ \citep{bie06}.
334:
335: We averaged our polarization data obtained on 46 nights when R CrB was
336: at maximum brightness stage ($\Delta m_{\rm V} \leq 0.3$ mag)
337: except those in the periods of the transient polarization.
338: The mean values for Stokes $Q$ and $U$ parameters and their standard
339: deviations in each band are
340: $(Q_{\rm B}, U_{\rm B}) = ( -0.198\pm 0.068\ \% , -0.020\pm 0.055\ \% )$,
341: $(Q_{\rm V}, U_{\rm V}) = ( -0.203\pm 0.049\ \% , -0.035\pm 0.053\ \% )$,
342: $(Q_{\rm R}, U_{\rm R}) = ( -0.188\pm 0.044\ \% , -0.028\pm 0.056\ \% )$, and
343: $(Q_{\rm I}, U_{\rm I}) = ( -0.161\pm 0.044\ \% , -0.026\pm 0.055\ \% )$.
344: Parameters of Serkowski function \citep{ser75,whi92b} corresponding
345: to those values are derived by a non-linear regression as
346: $p_{\rm max}=0.204\pm 0.046$ \%,
347: $\lambda_{\rm max}=0.48\pm 0.27$ $\micron$ and
348: $\theta=94\fdg 1 \pm 4\fdg 1$.
349: These are consistent with the constant and/or interstellar
350: component derived in previous studies.
351: The constancy of the component over two decades, being much more
352: than the $\sim$3.5 yr IR period, allows us to
353: consider that the component is fully interstellar polarization.
354: %although it is hard to know whether
355: %the component is fully interstellar polarization or partially
356: %originated in dust clouds associated with R CrB itself
357: %\citep{gil86,jef95,ohn01,ohn03,yud02,yud03,lav04,lea07}.
358: It is noted that \citet{cla95} found a small variability of
359: polarization up to 0.14 \%\ (possibly correlated with the light curve,
360: at least, for their 1991 data) during the maximum brightness stage.
361: The variability seems rather random, and such an effect would be
362: cancelled out in our averaged values, while they possibly remain
363: in the standard deviations.
364:
365:
366: \section{DISCUSSION}
367:
368: After vectorial subtraction of the constant component from the
369: observations, we obtain variable components of polarization,
370: $\ p_{\rm var}(\lambda )$ and $\theta_{\rm var}(\lambda)$.
371: Their values in synthetic $V$ band are shown in Table \ref{tbl-1}.
372:
373: It is known that polarized flux of R CrB is relatively constant
374: since 1968 at the level of $\sim 10^{-3}$ to $10^{-4}\ F_{\rm *}$
375: at any part of declines \citep{whi92a,cla95,cla97},
376: where $F_{\rm *}$ is the total stellar flux at maximum brightness stage.
377: Here, the polarized flux is simply derived as
378: \begin{equation}
379: f_{\rm pol} = p_{\rm var} \times 10^{-0.4 \Delta m_{\rm V}}
380: \ F_{\rm *}\mbox{ ,}
381: \end{equation}
382: where $\Delta m_{\rm V}$ is the drop of visual
383: magnitude from maximum $V_{\rm max}=6.09$ (see \S~1).
384: In our data (Figure \ref{fig1}d), $f_{\rm pol}$ also falls between
385: $10^{-3}$ and $10^{-4}\ F_{\rm *}$ on almost all days both during
386: and out of declines.
387: However, at the epochs of the transient polarization,
388: $f_{\rm pol}$ showed $\sim$5 times as much as the typical
389: maximum values, which indicates that the net polarization
390: flux was extraordinarily large.
391:
392:
393: \subsection{Origin of the Transient Polarization}
394:
395: Our polarimetric monitoring brought first detection of the
396: `transient polarization' amounting to 0.5 \% for R CrB in
397: the maximum brightness phase.
398: As far as we know, there exists only one observation in which
399: R CrB near maximum brightness showed polarimetric variation larger
400: than $0.2$ \%: \citet{efi80} found $\Delta p = 0.38\pm 0.03$ \%\
401: at $\lambda_{\rm eff}=$5450 \AA\ on 1974 August 24 $=$
402: JD$2,442,284$.
403: Figures \ref{fig5}a and \ref{fig5}b show the time variation of
404: $p_{\rm var}(\lambda)$ and $\theta_{\rm var}(\lambda)$
405: in the periods including the first and second events of the
406: transient polarization.
407: There is no significant wavelength dependence
408: in position angle for both events, compared with
409: the polarization found during decline phase
410: (Figures \ref{fig6}a--c).
411:
412: Suppose that several clouds exist around R CrB and that they
413: scatter the light from the star.
414: If each cloud was ejected from the photosphere in a random manner
415: and it contains a different polarizing property from one to another,
416: the position angle of the scattered light would be mostly
417: wavelength dependent.
418: The nearly constant $\theta$ against $\lambda$ indicates
419: that the polarization can be `effectively' attributed
420: to scattering by one component.
421: The wavelength dependence of $p_{\rm var}(\lambda)$ was
422: almost flat in optical wavelengths
423: and quite similar between the two cases.
424: For the event on 1974 August 24, $p(\lambda )$
425: seems also flat (0.38--0.46 \%) in the wavelength range,
426: 3630--7450 \AA\ \citep{efi80}.
427: The similarity in $p_{\rm var}(\lambda )$ curve suggests
428: an existence of the possible generality in the polarizing mechanism
429: among those three events.
430: Rayleigh scattering by large molecules or smallest solid particles
431: ($\lesssim \lambda / 10$) cannot reproduce such wavelength dependence
432: because the scattering coefficient is highly wavelength-dependent
433: ($\propto \lambda^{-4}$).
434: Scattering by free electrons can reproduce a flat
435: $p_{\rm var}(\lambda )$ curve.
436: However, it is unlikely for R CrB to have a large amount of
437: ionized gas which can produce an observable polarization
438: because the effective temperature of R CrB is not so high
439: ($T_{\rm eff}\simeq 6900$ K; \citealt{sch75,asp97}).
440: The emission lines which appears in R CrB during declines
441: (e.g., \citealt{rao99,rao06}) are not so strong that spectra
442: at maximum brightness stage have little emission line.
443:
444: As described in \S 1, several observational facts suggest
445: that R CrB intermittently ejects dust puffs even at
446: maximum brightness phases.
447: The transient polarization could be naturally explained by a
448: dust puff ejection off the line of sight.
449: The shallow dips in the lightcurve (Figure \ref{fig2}) would
450: support this idea, if we can assume that the dust puff eclipses
451: a small part of the photosphere at the very early (nearest) stage.
452: Single-sized grain of radius $a \simeq 0.1\ \micron$ shows a peak
453: polarization at optical wavelength and the peak wavelength
454: varies with the grain radius (e.g., \citealt{sha75}).
455: For flat $p_{\rm var}(\lambda)$ curve, the dust grains would have
456: a size distribution of a wide range of $a$.
457:
458: We performed a calculation of Mie scattering with a simplified
459: model consisting of a point-like dust cloud and a central star,
460: and estimated the light from the stellar system; that is the sum
461: of an unpolarized direct stellar light and the
462: partially-polarized scattered light.
463: In the calculation we assumed that the dust grains are all
464: spherical amorphous carbon (e.g., \citealt{lam01}) and adopted
465: the refractive indices of the BE soot \citep{rou91}.
466: We also assumed that the cloud is optically-thin, i.e., multiple
467: scattering process is negligible.
468: We set the distance of the cloud $r=2R_{\rm *}$ where $R_{\rm *}$
469: is the stellar radius ($= 85 R_{\sun}$; \citealt{fea75}) and
470: make an estimation of the total mass of dust necessary for
471: explaining the $0.5$ \% polarization.
472: %The level of the polarization amplitude is scaled by the total
473: %mass of dust without notable change in the curvature of $p(\lambda )$,
474: %but the error in the the estimated mass becomes significant for larger
475: %mass because we have not taken into account the multiple scattering
476: %effects. (For more realistic estimation of the dust mass, see Appendix.)
477: The results are shown in Figure \ref{fig7}.
478: In the case of forward-scattering (scattering angle
479: $\alpha < 90\arcdeg$, where $\alpha$ is the angle between
480: the direction of the incident light and that of the
481: scattered one),
482: the $p(\lambda )$ curves gradually decrease with wavelength
483: and they do not fit well the observation.
484: For a right angle scattering and backward-scattering
485: ($\alpha > 90\arcdeg$) cases,
486: the single-sized grain model fails to reproduce
487: the flat $p(\lambda)$ curve. The size distribution
488: of the carbon dust should range over $\sim$0.13 $\micron$
489: in order to produce the flat $p_{\rm var}(\lambda)$ curve.
490: We here consider a power-law size distribution
491: like a well-known interstellar dust model
492: (\citealt{mat77}; hereafter, MRN).
493: The MRN size distribution is expressed as $n(a)\propto a^{-\beta}$
494: within $a_{\rm min} \leq a \leq a_{\rm max}$, where
495: $\beta = 3.5,\ a_{\rm min}=0.005\ \micron$ and
496: $a_{\rm max}= 0.25\ \micron$.
497: Results of the MRN model seem not to contradict
498: the observation in a wide range of scattering
499: angle ($\alpha=60\arcdeg$--$135\arcdeg$).
500: This is only an illustrative example.
501: A large choice of parameters
502: ($\alpha,\ a_{\rm min},\ a_{\rm max},\ \beta$) can yield
503: similar $p_{\rm var}(\lambda )$ curves.
504: In general, the scattering by grains having
505: various sizes seems a plausible mechanism for the transient
506: polarization.
507: The similarity in the observed $p(\lambda )$ curves between/among
508: two (or three, including \citealt{efi80}'s event) events suggest
509: that the dependence of $p(\lambda )$ on the scattering angle
510: is not so strong.
511:
512: A correlation between pulsational phase and the time of
513: decline onset has been found for RY Sgr and V854 Cen which show
514: fairly regular pulsation cycles than R CrB \citep{pug77,law92}.
515: Although the correlation is less significant for R CrB itself
516: \citep{pug77,gon83,per87,law91,fer94}, it is recently confirmed
517: by \citet{cra07}.
518: These suggest that the condensation of dust is triggered
519: by the stellar pulsation.
520: According to this picture, the appearance of the transient
521: polarization is likely to correlate with the pulsation phase.
522: In our observation, the interval between the two polarization
523: events was $151\pm 1$ d.
524: It has been known that R CrB shows a semi-regular pulsation
525: of about 40 d from photometric and spectroscopic
526: observations, for example $38.6$ d \citep{ale72},
527: $P=39.8$ d \citep{yud02} and $P=42.7$ d \citep{rao99}.
528: If we assume that the 151 d interval corresponds to four
529: cycles, the length of one cycle is $\sim 38$ d.
530: This seems a little shorter compared with the pulsation period.
531: However, it is noted again that the pulsation period of R CrB is
532: not coherent. Some shorter periods have been found in a timescale of
533: a few hundred days.
534: \citet{fer94} found well-developed oscillations with an
535: apparent period of $36.7\pm 0.9$ d in late 1992 in the visual
536: light curve; while they also found that the better-defined
537: peaks observed from 1992 through 1993 could be fitted with
538: a single, linear ephemeris of period $35.3\pm 0.2$ d.
539: We cannot exclude the possibility that the transient
540: polarization was correlated with the stellar pulsation.
541:
542:
543: \subsection{Size Distribution of Dust}
544:
545: In the previous section, we showed that the wavelength dependence
546: of the transient polarization can be reproduced by light scattering
547: due to the carbon grain with various sizes including grains larger
548: than $\sim$0.13 \micron.
549: However, UV extinction properties at the 1980 decline of R CrB
550: indicate that the dust was grassy or amorphous carbon grain
551: with a size distribution of $a=0.005$--$0.06\ \micron$ \citep{hec84}.
552: A detailed analysis of extinction curves \citep{zub97}
553: for three observations of R CrB during declines in 1980--1984
554: indicated that the size distribution of graphite or amorphous
555: carbon grains has a distinct peak-like form (widths of
556: $\Delta a=0.002$--$0.006\ \micron$) with typical sizes
557: of $a = 0.03$ through $0.07\ \micron$.
558: If such small particles scatter the light from the central star,
559: the $p_{\rm var}(\lambda)$ curve should show a slope of
560: decreasing with wavelength in the optical (Figure \ref{fig7}).
561:
562: The amplitude of polarization is proportional to inverse square
563: of the distance between the central star and the dust puff, $r$,
564: while the extinction depends only on the column density
565: along the line of sight.
566: The information derived from the extinction would be
567: a mean property of the dust clouds along the line of sight
568: toward the star, while the information
569: derived from the polarization would be selectively for the dust
570: close to the central star.
571: Therefore the apparent difference in dust size between
572: the estimate from polarimetry and that from UV extinction
573: may indicate that the size distribution of the dust changes
574: with the distance from the central star.
575: \citet{coy73} reported the variability of the wavelength
576: dependence of polarization in 1972 decline of R CrB, and
577: interpreted each observation with scattering by graphite dust
578: of different sizes ranging from $0.05$ to $0.10\micron$ in radius.
579: In their results the mean particle size became smaller
580: unsteadily with time.
581: \citet{eva85} proposed that the extinction
582: properties of the dust responsible for 1982 decline of the
583: hot RCB star, MV Sgr, can be explained by somewhat large
584: carbon grains ($a\sim 0.2\ \micron$).
585: Another model applied for the same extinction data also
586: gave a similar result \citep{zub97}.
587: For some extinction data of other RCB stars, he also suggested
588: that the particle size exceeds $a=0.1\ \micron$.
589: These facts suggest that the scattering matter consist of
590: discrete clouds with different mean particle sizes
591: (depending on their ages and their distance from the star)
592: even in the same star.
593: If the dust forms near the photosphere,
594: a grain evaporation mechanism might play a role on
595: the change of the size distribution (e.g., \citealt{eva93}).
596: The particle size inferred from our polarimetry
597: suggests that the evolutionary phase of the dust grains
598: causing the transient polarization is more or less the same
599: from that of the dust grains obscuring the stellar flux
600: during deep declines.
601:
602:
603: %Both authors mentioned that differences in the grain size
604: %can occur among independent extinction episodes
605: %in the same object.
606: %Therefore, we cannot exclude the possibility that
607: %the dust puff causing the transient polarization
608: %actually contains larger grains.
609:
610:
611: \subsection{On the Short Timescale of Fading}
612:
613: The transient polarization in 2001 March
614: diminished gradually between March 11 and 13.
615: For the second event, the polarization also diminished within two days.
616: It is remarkable that the decaying timescale of the transient
617: polarization is only $\sim 2$ days, which is much shorter
618: than the recovery times of the brightness from deep minimum
619: (typically several weeks).
620: If we assume that the scattering geometry keeps similar,
621: the polarized flux approximately depends on the product of
622: intensity of the stellar light at the scattering material
623: and total scattering cross section.
624: Thus, the amplitude of polarization decreases as the dust cloud
625: goes away from the star (i.e., larger $r$) or the dust particles
626: are destroyed (i.e., smaller $a$).
627:
628: Here we consider the former case. If the decrease of polarization
629: was only due to radial dispersal of dust cloud, the rate of
630: decrease in polarization roughly depends on the radial distance $r$
631: and the mean expansion velocity of the cloud $v_{\rm exp}$.
632: The distance can be expressed as $r=R_{\rm df}+v_{\rm exp}t$,
633: where $R_{\rm df}$ is the distance between the central star
634: and the place where the dust formation finishes, and $t$ is
635: the time elapsed from the cessation of the dust formation.
636: For simplicity, we assume that the scattering cloud is optically thin.
637: In this case, the polarization amplitude approximately changes as
638: $p(r) = p_{\rm c} D(r)(R_{\rm *}/r)^{2}$,
639: where $p_{\rm c}$ is a constant and
640: $D(r)=\{1-(R_{\rm *}/r)^{2}\}^{0.5}$ is the
641: finite disk depolarization factor \citep{cas87}.\footnote{The
642: optical thickness of the cloud might not be
643: practically negligible, which results in depolarization.
644: As the cloud expands, the cloud becomes
645: optically-thin and the polarizing efficiency gradually recovers.
646: Thus, $p$ should be damped more slowly than $D(r)\cdot r^{-2}$ in the
647: vicinity of the star and we obtain the lower-limit of
648: $v_{\rm exp}$ with this method anyhow.}
649: This function monotonically decreases with $r$ for
650: $r>\sqrt{3/2}R_{\rm *}$, and the most rapidly decreasing
651: case occurs for nearly $R_{\rm df}=\sqrt{3/2}R_{\rm *}$.
652: The observed transient polarization decreased to
653: one fourth within $\sim 2$ days.
654: These indicate that the dust cloud needs to travel,
655: at least, $1.9 R_{\rm *}$ within two days, i.e.,
656: $v_{\rm exp} \gtrsim 650$ km s$^{-1}$.
657: Thus, if the expansion velocity exceeds $650$ km s$^{-1}$,
658: the dispersal of dust cloud can explain the timescale
659: of the polarization extinction.
660: However, the derived velocity is not consistent with the
661: typical mass loss velocity observed in R CrB.
662: The absorption component of the \ion{Na}{1} D or the
663: \ion{He}{1} $\lambda$10830 line are blueshifted by up to
664: $-150$ to $-350$ km s$^{-1}$
665: (e.g., \citealt{pay63,que78,cot90,rao99,cla03}).
666: Evolutions of emission line system at the onset of
667: declines suggest that the wind velocity is likely to be lower
668: ($\sim 50$ km s$^{-1}$) at close proximity to
669: the photosphere, $\simeq 1$--$2 R_{\rm *}$
670: \citep{pay63,ale72,cla96}.
671: Although higher-velocity wind have been reported in
672: the unusual RCB stars, V854 Cen (400--1000 km s$^{-1}$,
673: \citealt{cla93,rao93}), the observational facts
674: suggests the ejecta of R CrB is gradually
675: accelerated up to $\sim 200$ km s$^{-1}$ within a
676: few weeks (probably by radiation pressure on dust grains).
677: It is unlikely that the dust puff is accelerated
678: to $v_{\rm exp} \gtrsim 650$ km s$^{-1}$ within
679: $R\sim 1.2R_{\rm *}$.
680:
681: %\citet{cla93} carried out low-dispersion UV spectroscopy of
682: %V854 Cen and found emission line profile variation implying
683: %gas velocities of $\gtrsim 1000$ km s$^{-1}$ from.
684: %Optical observations for the star also showed an existence of
685: %component having velocity of
686: %$\simeq$400--500 km s$^{-1}$ in \ion{Na}{1} D absorption
687: %line \citep{rao93,cla93}.
688: %%%%
689:
690: Therefore, the rate of depolarization can hardly be explained
691: only by the radial expansion of the dust puff.
692: If the dust forms actually near the photosphere
693: ($\lesssim 2 R_{\rm *}$), the strong radiation from the
694: central star can evaporate and destroy the dust grains once the
695: possible cooling instability to promote the dust condensation
696: (see \S 1) terminates.
697: In any case, the rapid disappearance of polarization
698: prefers that the dust grains are formed in an activated region,
699: that is, a region close to the photosphere.
700: %To give a more reliable explanation, simultaneous
701: %observations (e.g., optical high-dispersion spectroscopy
702: %and/or near-IR photometry and spectroscopy) as well as
703: %further spectropolarimetry will be required.
704:
705: The problem of dust formation in such hostile environments is
706: also under discussion in the field of Wolf-Rayet stars.
707: Some phenomenological similarities
708: with RCB stars have been reported for some late-type WC
709: stars and [WC] PN nuclei (e.g., \citealt{wil97,cro97,vee98,cla01,kat02}).
710: %The role of UV photons in dust formation is discussed in
711: %the field of dusty classical novae \citep{sho04}.
712: Polarization is relatively sensitive to the region close
713: to the light source for some optically-thin systems,
714: and the transient polarization events represent
715: the potential of polarimetry as the probe to the
716: dust formation around such stars.
717:
718:
719: \subsection{Geometry of Mass Loss}
720:
721: \citet{sta88} presented that the PA of polarization in R CrB
722: observed during decline of 1986 was along with data of two
723: previous declines of 1968 and 1972 \citep{ser69,coy73}.
724: This suggests that there was a preferred plane
725: (PA $\simeq 30$--$45\arcdeg$) in the dust ejection.
726: However, the PAs observed during 1989 and 1993 declines
727: were significantly different from them \citep{cla95}.
728: \citet{cla97} summarized a data set in five declines
729: with polarimetry (from 1971 through 1996) having wavelength
730: information when R CrB was more than 5 mag below maximum
731: brightness, and presented that they well fit to the
732: two opposite sector regions of PA $=15\arcdeg\pm 15\arcdeg$ and
733: $=105\arcdeg\pm 15\arcdeg$ in $QU$ diagram.
734: They suggested that dust ejection in R CrB has
735: a bipolar geometry consisting of obscuring torus
736: and bipolar lobes.
737: We can see a similar tendency in our data at decline
738: phases (Figure \ref{fig8}).
739: The PA of the largest polarization found during
740: the 2000 decline ($90\arcdeg$--$100\arcdeg$) falls
741: in the sector regions.
742: The PA in the recovering phase of the 2003 decline
743: ($30\arcdeg$--$36\arcdeg$) also falls near the sector regions.
744: These implies that the bipolar geometry for dust ejection
745: directions still exists in R CrB.
746: It is noted that the transient polarization was at
747: $80\arcdeg$--$90\arcdeg$ (2001 March) and $\sim 70\arcdeg$
748: (2001 August), which lies near or only slightly apart from
749: the edge of the sector regions.
750: %$30\arcdeg$--$50\arcdeg$ in 1999 December
751: %and $90\arcdeg$--$100\arcdeg$ in 2000 December, respectively.
752: %Across the 2003 decline the PA shows an excursion from $\sim 140\arcdeg$
753: %to $\sim 40\arcdeg$.
754:
755: During deep declines a considerable rotation ($\gtrsim 40\arcdeg$)
756: of the PA with wavelength has been occasionally found in
757: some RCB stars (e.g., \citealt{whi92a,cla97}).
758: This can be explained by an optical depth effect where a
759: thick torus obscures the central star and thin bipolar
760: lobes act as a reflection nebula \citep{cla97}.
761: %suggests that there are, at least, two clouds at different
762: %PAs on the projected sky, and the clouds contains dust grains having
763: %different size distribution from each other.
764: %The eclipse of the unpolarized direct light relatively
765: %emphasizes the polarization due to scattering by dust, and
766: %more distant clouds, which has been ejected earlier,
767: %can relatively contribute to the observable polarization.
768: %Assuming that the difference of the size distribution
769: %reflects the age of the cloud, the rotation of the position
770: %angle might trace temporal changes in the direction of dust
771: %ejection just before the decline.
772: On the other hand, for the transient polarization,
773: the rotation of the PA with wavelength is small.
774: It supports the idea that only a single dust puff, which is
775: closest to the photosphere, contributes the transient
776: polarization effectively. The size distribution discussed
777: in \S 4.2 would reflect the real properties of the newborn
778: grains selectively.
779:
780: %As mentioned in \S 1, it has long been considered that
781: %R CrB ejects a dust puff toward a random direction.
782: \citet{ohn01} carried out {\it K}-band speckle interferometric
783: observations of R CrB on 1996 October 1 ($V=7$) and 1999 September
784: 28 ($V=10.61$) and suggested that the large dust shell
785: (with an inner boundary at $r=82 R_{\rm *}$ and 920 K) has a
786: rather isotropic distribution, at least, in a scale of larger
787: than $\sim 300 R_{\rm *}$.
788: The authors also introduced a newly formed dust cloud close to
789: the star ($r=4.5 R_{\rm *}$ and 1200 K) to fit both the visibility
790: and the spectral energy distribution simultaneously.
791: \citet{asp97} gave an analysis of SED with a model atmosphere and
792: suggested that the IR excess could be explained by recently formed,
793: hot dust grains ($\sim 2000$K) amounting to a total mass of
794: $\sim 10^{-11}\ M_{\sun}$.
795: Some extinction models for RCB stars suggested that
796: the total mass (gas$+$dust) per ejection event were
797: $10^{-9}$--$10^{-7}\ M_{\sun}$ \citep{fea86,fad88,cla92}
798: although these could be affected by uncertainties in $r$,
799: $\phi$ (opening angle of the dust puff), and the gas-to-dust ratio.
800: \citet{yud02} derived a mean dust formation rate of
801: $3.1\times 10^{-9}\ M_{\sun}$ yr$^{-1}$ with a radiative
802: transfer modeling.
803: Our Monte Carlo simulation (see Appendix) suggests that
804: the dust puff off the line of sight can produce a polarization
805: of $\simeq 0.5$\% when it has a mass of
806: $\gtrsim 9\times 10^{-13}\ M_{\sun}$ and an opening angle of
807: $\gtrsim 40\arcdeg$ for $r\sim 2 R_{\rm *}$.
808: The actual mass should be highly model dependent.
809: However, the lower-limit of the dust mass is still much less than
810: the mean dust formation rate ($\sim 10^{-9}\ M_{\sun}$ yr$^{-1}$),
811: and we conclude that the amplitude of polarization in the
812: transient polarization events would be consistent with
813: the random ejection model in which dust puffs are formed
814: in the close vicinity of the stellar atmosphere.
815:
816:
817: \section{SUMMARY AND CONCLUSION}
818:
819: We carried out spectropolarimetric monitoring of R CrB
820: from 1998 January to 2003 September.
821: The constant component of polarization derived by \citet{sta88}
822: and confirmed by \citet{cla95} still existed in our observational period.
823: In addition, we discovered a temporal increase of polarization up to
824: $\simeq 0.5$ \% at maximum brightness stage in 2001 March and August.
825: It is likely that dust puff which was ejected off the line of sight
826: produced the transient polarization.
827: Shallow declines ($\sim$0.1 mag) were marginally found a few days
828: prior to the polarization events, which would support this idea
829: if we can assume that the dust puff eclipsed a part of the photosphere
830: at the very early stage.
831: The flatness in $p_{\rm var}(\lambda)$ curve of the transient
832: polarization, which was a common property for both two cases,
833: suggested that the size distribution of the dust was
834: not like a $\delta$-function (as the results of UV--optical
835: extinction studies) but rather wide, e.g., a well-known
836: model of interstellar dust (MRN size distribution).
837: We suggest that the size distribution evolves
838: with the age or the distance from the star.
839: The fading timescale ($\sim 2\pm 0.5$ days) of the polarization
840: is fairly rapid compared with the recovering rate of brightness from
841: deep declines.
842: It cannot be interpreted by a simple dissipation
843: due to radial expansion of the dust puff at the
844: expansion velocity of $\sim 200-350$ km s$^{-1}$,
845: even in the case where the dust formation
846: occurred at $r \sim 2\ R_{\rm *}$.
847: Higher expansion velocity or some mechanism to
848: destroy the dust grains should be needed.
849:
850: So far we have not detected a possible increasing
851: part of the transient polarization, which would be
852: important to discuss the process and the site of
853: the dust formation around RCB stars.
854: Further spectropolarimetric monitoring of RCB stars near
855: maximum brightness will bring a valuable information of a
856: dust formation and destruction processes in a hostile
857: circumstellar environment.
858:
859:
860: \acknowledgments
861:
862: We thank G.~C.~Clayton, the referee, for helpful suggestions.
863: We are grateful to T.~Cho,
864: H.~Iwamatsu, S.~Hamasaka, N.~Hirakata, R.~Hirata,
865: K.~Homma, K.~Hoshino, T.~Karube, N.~Kobayashi,
866: M.~Kondoh, S.~Masuda, Y.~Nakamura, S.~Nakayama,
867: A.~Okazaki, I.~Ota, M.~Saito, T.~Sakurai, R.~Suzuki,
868: C.~Tokoku, K.~Yoshioka
869: for their kind supports to the observation.
870: We also thank to the staff members of Dodaira Observatory
871: and Okayama Astrophysical Observatory of NAOJ for technical
872: supports, especially to Y. Norimoto, H. Shibasaki and K. Okita.
873: We are indebted to the variable star observations from the
874: VSNET and AAVSO database contributed by observers worldwide
875: for the light curve of R CrB.
876: Data reduction/analysis was in part carried out on the ``sb'' computer
877: system operated by the Astronomical Data Analysis Center (ADAC)
878: and Subaru telescope of NAOJ.
879: NAOJ is an interuniversity research institute of astronomy
880: operated by the Ministry of Education, Culture, Sports, Science and
881: Technology.
882: This work was supported by a Grant-in-Aid from the Ministry
883: of Education, Culture, Sports, Science and Technology of
884: Japan (No. 17684004).
885:
886: \appendix
887:
888: \section{MONTE-CARLO SIMULATION OF POLARIZATION BY DUST PUFF}
889:
890: Monte Carlo method is a useful tool to solve
891: problems including multiple scattering of light
892: since it requires only a few assumptions and it can be applied
893: for many kinds of complicated geometries and particles.
894: The details in application of Monte Carlo method for
895: light scattering process is described in, e.g.,
896: \citet{war83}, and \citet{hil91}.
897:
898: As seen in Figure \ref{fig9},
899: we consider the Cartesian coordinate, (x, y, z).
900: We define that the origin is the center of the star.
901: The stellar radius is 1, i.e., the coordinate is normalized
902: by the stellar radius.
903: The direction of the earth is {\boldmath $e$}$=(0,0,1)$.
904: The shape of the dust puff is a part of geometrically-thin spherical shell
905: with a radius of $r$ and an opening angle of $\phi$ from the origin.
906: The center of the dust puff, {\boldmath $r$}, is located
907: in the $x$-$z$ plane.
908: These enable us to investigate the depolarization effect
909: due to finite solid angle of the photosphere in the neighborhood of the
910: star, which is a different point from \citet{cod95}'s.
911: Optical depth of the cloud along {\boldmath $r$} is $\tau _{0}$.
912: For simplicity, we neglect geometrical thickness of the shell.
913: The scattering angle, that is, the angle between
914: {\boldmath $r$} and {\boldmath $e$}, is $\alpha$.
915:
916: In our Monte Carlo simulation,
917: more than a million photons are emitted from random sites
918: on the stellar surface toward random directions.
919: All released photons have the same intensity,
920: which means that rim darkening effect is neglected.
921: When they meet the cloud, some penetrate,
922: some are absorbed by, and the others are scattered by
923: dust grains in the cloud.
924: The scattered photons are partially polarized and emitted again
925: toward a new direction according to a phase function
926: {\boldmath $R$}$(\alpha )$.
927: We adopt a single peaked Henyey-Greenstein function \citep{whi79}
928: as the phase function:
929: \begin{equation}
930: \mbox{\boldmath $R$} (\alpha) = \frac{3}{4} \left(
931: \begin{array}{cccc}
932: P_1 & P_2 & 0 & 0 \\
933: P_2 & P_1 & 0 & 0 \\
934: 0 & 0 & P_3 & -P_4 \\
935: 0 & 0 & P_4 & P_3
936: \end{array}
937: \right),
938: \end{equation}
939: and,
940: \begin{eqnarray}
941: P_1 &=& \frac{1-g^2}{(1+g^2-2g\cos \alpha)^{3/2}}, \\
942: P_2 &=& -p_{\rm l} \frac{1-\cos ^2 \alpha}{1+\cos ^2 \alpha} P_{1}, \\
943: P_3 &=& \frac{2\cos \alpha}{1+\cos ^2 \alpha} P_{1}, \\
944: P_4 &=& -p_{\rm c}\frac{1-\cos ^2 \alpha _{\rm f}}{1+\cos ^2
945: \alpha _{\rm f}} P_{1}, \\
946: \alpha _{\rm f} &=& \alpha (1+3.13s \exp [-7\alpha / \pi]),
947: \end{eqnarray}
948: where $g$ is an asymmetric parameter
949: (ranging from 0 for isotropic scattering to 1 for forward-throwing),
950: $p_{\rm l}$ and $p_{\rm c}$ are the maximum linear and circular
951: polarization, respectively,
952: and $s$ is the skew factor ($=1$ in almost all cases).
953: Three parameters $\alpha$, $g$,
954: and $p_{\rm l}$, are related with the linear polarization.
955: $g$ varies with a wavelength and a mean size of dust.
956: \citet{whi79} investigated the dependency of $g$ on wavelength
957: for a well-known model of interstellar graphite dust.
958: We used his values, lacking a better alternative for
959: a mixture of cosmic grains having various sizes.
960: Referring to Figure 2 of his paper, we adopt $g=0.46$ at 5500 \AA.
961: We also obtain the albedo $\omega = 0.61$ at 5500 \AA\ from
962: the same figure.
963: The $p_{\rm l}$ is set to 0.51 \citep{cod95}.
964: Although the single peaked Henyey-Greenstein phase function is
965: not well approximated at longer wavelength region and for
966: back-scattering cases \citep{whi79},
967: the difference can be negligible in our cases,
968: i.e., $x = 2\pi a / \lambda
969: \lesssim 1$ and $\alpha \leq 135\arcdeg$.
970: More details of our calculation will be described in a
971: preparing paper \citep{ike07}.
972:
973:
974: \subsection{Size and Mass of Dust Puff}
975:
976: In this section, we give a rough estimation for the size
977: and mass of dust puff responsible for the amplitude of the
978: transient polarization ($p_{\rm var}\simeq 0.5$\%)
979: using the Monte-Carlo method.
980: In Table \ref{tbl-2} we summarized the results at 5500 \AA\ for both
981: $r=2 R_{\rm *}$ and $20 R_{\rm *}$ cases.
982: Since the wavelength dependence is expected to be small
983: for the dust model, they can be considered as typical values
984: in the optical region.
985: It is obvious that the models of $\alpha' = 135 ^{\circ}$
986: (backward scattering) cannot produce $p > 0.5$,
987: where $\alpha'$ is the scattering angle at the cloud center.
988: Figure \ref{fig10} shows the relations of $\tau _{0}$ vs. $\phi$
989: satisfying polarization of $0.5$\%.
990: To obtain the results of $p \simeq 0.5$ \%, we need (i)
991: $\tau_{0}\gtrsim 0.5$--$1$ and $\phi \gtrsim 40\arcdeg$ for
992: $r=2 R_{\rm *}$, or (ii) $\tau_{0}\gtrsim 0.5$ and
993: $\phi \gtrsim 20\arcdeg$ for $r=20 R_{\rm *}$.
994: The cloud at larger distance can produce $p=0.5$ \% with smaller
995: optical thickness and smaller opening angle because the
996: depolarization effect due to finite solid angle of the
997: photosphere become small.
998: However, larger mass is needed for large $r$.
999: We can estimate the total mass of the dust puff as
1000: \begin{eqnarray}
1001: M_{\rm d} & = & \frac{4\pi}{3}\rho \langle a^3\rangle
1002: \frac{\tau _{0}}{\langle \sigma _{\rm ext}\rangle }
1003: r^2 \Omega \\
1004: &\simeq & 2.7 \times 10^{-12}
1005: \left( \frac{\tau _{0}}{1.0} \right)
1006: \left( \frac{r}{R_{\rm *}} \right)^2
1007: \left( 1 - \cos \frac{\phi}{2} \right) \quad M_{\sun},
1008: \label{eqn2}
1009: \end{eqnarray}
1010: where $\rho$ is the bulk density of dust grains,
1011: $\langle a^3\rangle$ and $\langle \sigma _{\rm ext}\rangle$ are
1012: the averaged cubic radius and averaged cross section of extinction
1013: per a dust particle, respectively, and $\Omega$ is solid angle
1014: of the puff seen from the star.
1015: In Equation \ref{eqn2} we adopted
1016: $\langle a^3\rangle = 3.80 \times 10^{-18}$ cm$^{3}$ and
1017: $\langle \sigma _{\rm ext}\rangle =1.42 \times 10^{-12}$
1018: cm$^{2}$ \citep{whi79},
1019: $\rho=2.2$ g cm$^{-3}$ \citep{eva93}.
1020: We have
1021: $M_{\rm d} \simeq 9 \times 10^{-13}\tau_{0}\ M_{\sun}$
1022: for the $r=2 R_{\rm *}$ case ($\phi = 48^{\circ}$), and
1023: $3 \times 10^{-11}\tau_{0}\ M_{\sun}$ for the $r=20 R_{*}$
1024: case ($\phi = 28^{\circ}$).
1025:
1026:
1027: \begin{thebibliography}{}
1028: \bibitem[Alexander et al.(1972)]{ale72} Alexander, J. B.,
1029: Andrews, P. J., Catchpole, R. M., Feast, M. W.,
1030: Lloyd Evans, T., Menzies, J. W., Wisse, P. N. J., \& Wisse, M. 1972,
1031: \mnras, 158, 305
1032: \bibitem[Asplund \& Gustafsson(1996)]{asp96} Asplund, M., \& Gustafsson, B.
1033: 1996, in ASP Conf. Ser. Vol. 96, Hydrogen-Deficient Stars,
1034: eds. C. S. Jeffery and U. Heber (San Francisco: ASP), 39
1035: \bibitem[Asplund et al.(1997)]{asp97} Asplund, M., Gustafsson,, B.
1036: Kiselman, D., \& Eriksson, K. 1997, \aap, 318, 521
1037: \bibitem[Bessell(1990)]{bes90} Bessell, M. S. 1990, \pasp, 102, 1181
1038: \bibitem[Bieging et al.(2006)]{bie06} Bieging, J. H., Schmidt, G. D.,
1039: Smith, P. S., \& Oppenheimer, B. D. 2006, \apj, 639, 1053
1040: \bibitem[Cassinelli, Nordsieck, \& Murison(1987)]{cas87} Cassinelli,
1041: J. P., Nordsieck, K. H., \& Murison, M. A. 1987, \apj, 317, 290
1042: %\bibitem[Clayton \& Wickramasinghe(1976)]{cla76} Clayton, D. D., \&
1043: %Wickramasinghe, N. C. 1976, \apss, 42, 463
1044: \bibitem[Clayton et al.(1992)]{cla92} Clayton, G. C.,
1045: Whitney, B. A., Stanford, S. A., \& Drilling, J. S. 1992,
1046: \apj, 397, 652
1047: \bibitem[Clayton et al.(1993)]{cla93} Clayton, G. C.,
1048: Lawson, W. A., Whitney, B. A., \& Pollacco, D. L. 1993,
1049: \mnras, 264, L13
1050: \bibitem[Clayton et al.(1994)]{cla94} Clayton, G. C., Lawson, W. A.,
1051: Cottrell, P. L., Whitney, B. A., Stanford, S. A., \& de Ruyter, F. 1994,
1052: \apj, 432, 785
1053: \bibitem[Clayton et al.(1995)]{cla95} Clayton, G. C., Whitney, B. A.,
1054: Meade, M. R., Babler, B., Bjorkman, K. S., \& Nordsieck, K. H.
1055: 1995, \pasp, 107, 416
1056: \bibitem[Clayton(1996)]{cla96} Clayton, G. C. 1996, \pasp, 108, 225
1057: \bibitem[Clayton et al.(1997)]{cla97} Clayton, G. C., Bjorkman, K. S.,
1058: Nordsieck, K. H., \& Zellner, N. E. 1997, \apj, 476, 870
1059: \bibitem[Clayton(2001)]{cla01} Clayton, G. C. 2001, \apss, 275, 143
1060: \bibitem[Clayton, Geballe, \& Bianchi(2003)]{cla03} Clayton, G. C.,
1061: Geballe, T. R., \& Bianchi, L. 2003, \apj, 595, 412
1062: \bibitem[Clayton et al.(2007)]{cla07} Clayton, G. C.,
1063: Geballe, T. R., Herwig, F., Fryer, C., \& Asplund, M. 2007,
1064: \apj, 662, 1220
1065: \bibitem[Code \& Whitney(1995)]{cod95} Code, A. D., \&
1066: Whitney, B. A. 1995, \apj, 441, 400
1067: \bibitem[Cottrell, Lawson, \& Buchhorn(1990)]{cot90} Cottrell, P. L.,
1068: Lawson, W. A., \& Buchhorn, M. 1990, \mnras, 244, 149
1069: \bibitem[Coyne \& Shawl(1973)]{coy73} Coyne, G. V., \& Shawl, S. J.
1070: 1973, \apj, 186, 961
1071: \bibitem[Coyne(1974)]{coy74} Coyne, G. V. 1974, IBVS, No. 914
1072: \bibitem[Crause, Lawson, \& Henden(2007)]{cra07} Crause, L. A.,
1073: Lawson, W. A., \& Henden, A. A. 2007, \mnras, 375, 301
1074: \bibitem[Crowther(1997)]{cro97} Crowther, P. A. 1997, \mnras, 290, L59
1075: \bibitem[de Laverny \& M\'ekarnia(2004)]{lav04} de Laverny, P.,
1076: M\'ekarnia, D. 2004, \aap, 428, L13
1077: \bibitem[Efimov(1980)]{efi80} Efimov, Yu. S. 1980, Bull. Crimean
1078: Astrophys. Obs., 61, 88
1079: \bibitem[Efimov(1990)]{efi90} Efimov, Yu. S. 1990, AZh, 67, 494
1080: \bibitem[Evans(1993)]{eva93} Evans, A. 1993, The Dusty Universe
1081: (New York: Ellis Horwood)
1082: \bibitem[Evans(1985)]{eva85} Evans, A., Whittet, D. C. B., Davies, J. K.,
1083: Kilkenny, D., \& Bode, M. F. 1985, \mnras, 217, 767
1084: \bibitem[Fadeyev(1986)]{fad86} Fadeyev, Y. A. 1986,
1085: in Hydrogen Deficient Stars and Related Objects, eds. K. Hunger,
1086: D. Sch\"onberner, N. Kameswara Rao (Dordrecht: Reidel), 441
1087: \bibitem[Fadeyev(1988)]{fad88} Fadeyev, Y. A. 1988, \mnras, 233, 65
1088: \bibitem[Feast(1975)]{fea75} Feast, M. W. 1975, in
1089: IAU Symp. No. 67, Variable Stars and Stellar Evolution,
1090: eds. V. E. Sherwood and L. Plaut (Dordrecht: Reidel), 129
1091: \bibitem[Feast(1986)]{fea86} Feast, M. W. 1986,
1092: in Hydrogen Deficient Stars and Related Objects, eds. K. Hunger,
1093: D. Sch\"onberner, N. Kameswara Rao (Dordrecht: Reidel), 151
1094: \bibitem[Feast(1997)]{fea97a} Feast, M. W. 1997,
1095: \mnras, 285, 339
1096: \bibitem[Feast(2000)]{fea00} Feast, M. W. 2000, in IAU Symp. 177,
1097: The Carbon Star Phenomenon, ed. R. F. Wing (Dordrecht: Kluwer), 207
1098: \bibitem[Feast et al.(1997)]{fea97b} Feast, M. W.,
1099: Carter, B. S., Roberts, G., Marang, F., \& Catchpole, R. M. 1997,
1100: \mnras, 285, 317
1101: \bibitem[Feast \& Glass(1973)]{fea73} Feast, M. W., \&
1102: Glass, I. S. 1973, \mnras, 161, 293
1103: \bibitem[Fernie \& Seager(1994)]{fer94} Fernie, J. D., \&
1104: Seager,S. 1994, \pasp, 106, 1138
1105: \bibitem[Forrest, Gillett, \& Stein(1972)]{for72} Forrest, W. J.,
1106: Gillett, F. C., \& Stein, W. A. 1972, \apj, 178, L129
1107: %\bibitem[Gillet et al.(1986)]{gil86} Gillett, F. C., Backman, D. E.,
1108: %Beichman, C., \& Neugebauer, G. 1986, \apj, 310, 842
1109: \bibitem[Goncharova, Kovalchuk \& Pugach(1986)]{gon83} Goncharova,
1110: R. I., Kovalchuk, G. U., Pugach, A. F. 1983, Astrophysics, 19, 161
1111: \bibitem[Hecht et al.(1984)]{hec84} Hecht, J. H., Holm, A. V.,
1112: Donn, B., \& Wu, C. C. 1984, \apj, 280, 228
1113: \bibitem[Hillier(1991)]{hil91} Hillier, D. J. 1991, \aap, 247, 455
1114: \bibitem[Iben, Tutukov, \& Yungelson(1996)]{ibe96} Ibe, I. J.,
1115: Tutukov, A. V., \& Yungelson, L. R. 1996, \apj, 456, 750
1116: \bibitem[Ikeda \& Seki(2007)]{ike07} Ikeda, Y., \& Seki, M.
1117: 2007, in preparation.
1118: %\bibitem[Jeffery(1995)]{jef95} Jeffery, C. S. 1995, \aap, 299, 135
1119: \bibitem[Kato et al.(2002)]{kat02} Kato, T., Haseda, K.,
1120: Takamizawa, K., \& Yamaoka, H. 2002, \aap, 393, L69
1121: \bibitem[Kawabata et al.(1999)]{kaw99} Kawabata, K. S., et al. 1999,
1122: \pasp, 111, 898
1123: %\bibitem[Kholopov et al.(1988)]{kho88} Kholopov, P. N., et al. 1988,
1124: % General Catalogue of Variable Stars 4th Edition, (Moscow: Nauka
1125: %Publishing House)
1126: \bibitem[Lambert et al.(2001)]{lam01} Lambert, D. L., Rao, N. K.,
1127: Pandey, G., \& Ivans, I. I. 2001, \apj, 555, 925
1128: \bibitem[Lawson(1991)]{law91} Lawson, W. A. 1991, \mnras, 253, 625
1129: \bibitem[Lawson et al.(1992)]{law92} Lawson, W. A., Cottrell, P. L.,
1130: Gilmore, A. C., \& Kilmartin, P. M. 1992, \mnras, 256, 339
1131: %\bibitem[Le\~ao et al.(2007)]{lea07} Le\~ ao, I. C., de Laverny, P.,
1132: %Chesneau, O., M\'e karnia, D., \& De Medeiros, J. R. 2007, \aap, 466, L1
1133: \bibitem[Loreta(1934)]{lor34} Loreta, E. 1934, Astr. Nach., 254, 151
1134: \bibitem[Mathis, Rumpl, \& Nordsieck(1977)]{mat77} Mathis, J. S.,
1135: Rumpl, W., \& Nordsieck, K. H. 1977, \apj, 217, 425
1136: \bibitem[Ohnaka et al.(2001)]{ohn01} Ohnaka, K., et al. 2001, \aap, 380, 212
1137: \bibitem[Ohnaka et al.(2003)]{ohn03} Ohnaka, K., et al. 2003, \aap, 408, 553
1138: \bibitem[O'Keefe(1939)]{oke39} O'Keefe, J. A. 1939, \apj, 90, 294
1139: \bibitem[Payne-Gaposchkin(1963)]{pay63} Payne-Gaposchkin, C. 1960,
1140: \apj, 138, 320
1141: \bibitem[Percy et al.(1987)]{per87} Percy, J. R., Carriere, L. E. M.,
1142: \& Fabro, V. A. 1987, \aj, 93, 200
1143: \bibitem[Pugach(1977)]{pug77} Pugach, A. F. 1977, IBVS, 1277
1144: \bibitem[Querci \& Querci(1978)]{que78} Querci, M., \& Querci, F.
1145: 1978, \aap, 70, L45
1146: \bibitem[Rao et al.(1999)]{rao99} Rao, N. K.,
1147: et al. 1999, \mnras, 310, 717
1148: \bibitem[Rao \& Lambert(1993)]{rao93} Rao, N. K., \& Lambert, D. L.
1149: 1993, \aj, 105, 1915
1150: \bibitem[Rao, Lambert, \& Shetrone(2006)]{rao06} Rao, N. K.,
1151: Lambert, D. L., \& Shetrone, M. D. 2006, \mnras, 370, 941
1152: \bibitem[Rosenbush \& Rosenbush(1990)]{ros90} Rosenbush, A. \'E.
1153: \& Rosenbush, V. K. 1990, IBVS, 3439
1154: \bibitem[Rosenbush(2001)]{ros01} Rosenbush, A. \'E. 2001,
1155: Astrophysics, 44, 78
1156: \bibitem[Rouleau \& Martin(1991)]{rou91} Rouleau, F. \& Martin, P. G.,
1157: 1991, \apj, 377, 526
1158: \bibitem[Saio \& Jeffery(2002)]{sai02} Saio, H., \& Jeffery, C. S.
1159: 2002, \mnras, 333, 121
1160: \bibitem[Sch\"onberner(1975)]{sch75} Sch\"onberner, D. 1975,
1161: \aap, 44, 383
1162: \bibitem[Serkowski \& Kruszewski(1969)]{ser69} Serkowski, K.,
1163: \& Kruszewski, A. 1969, \apj, 155, L15
1164: \bibitem[Serkowski, Mathewson, \& Ford(1975)]{ser75} Serkowski, K.,
1165: Mathewson, D. S., \& Ford, V. L. 1975, \apj, 196, 261
1166: \bibitem[Shawl(1975)]{sha75} Shawl, S. J. 1975, \aj, 80, 595
1167: %\bibitem[Shore \& Gehrz(2004)]{sho04} Shore, S. N., \& Gehrz, R. D.
1168: %2004, \aap, 417, 695
1169: \bibitem[Soker \& Clayton(1999)]{sok99} Soker, N., \& Clayton, G. C.
1170: 1999, \mnras, 307, 993
1171: \bibitem[Stanford et al.(1988)]{sta88} Stanford, S. A.,
1172: Clayton, G. C., Meade, M. R., Nordsieck, K. H., Whitney, B. A.,
1173: Murison, M. A., Nook, M. A., \& Anderson, C. M. 1988, \apj, 325, L9
1174: \bibitem[Stein et al.(1969)]{ste69} Stein, W. A., Gaustad, J. E.,
1175: Gillett, F. C., \& Knacke, R. F. 1969, \apj, 155, L3
1176: \bibitem[Taylor(1984)]{tay84} Taylor, B. J. 1984, \apjs, 54, 259
1177: \bibitem[Trammell, Dinerstein, \& Goodrich(1994)]{tra94} Trammell,
1178: S. R., Dinerstein, H. L., \& Goodrich, R. W. 1994, \aj, 108, 984
1179: %\bibitem[Trimble \& Kundu(1997)]{tri97} Trimble, V., \& Kundu, A. 1997,
1180: %\pasp, 109, 1089
1181: \bibitem[Veen et al.(1998)]{vee98} Veen, P. M., van Genderen, A. M.,
1182: van der Hucht, K. A., Sterken, C., \& Dominik, C. 1998, \aap, 329, 199
1183: \bibitem[Warren-Smith(1983)]{war83} Warren-Smith, R. F. 1983,
1184: \mnras, 205, 337
1185: \bibitem[Wickramasinghe(1973)]{wic73} Wickramasinghe, N. C. 1973,
1186: Light Scattering Functions for Small Particles with Applications
1187: in Astronomy (London: Adam Hilger)
1188: \bibitem[White(1979)]{whi79} White, R.L. 1979, \apj, 229, 954
1189: \bibitem[Whitney et al.(1992)]{whi92a} Whitney, B. A., \& Clayton. G. C.,
1190: Schulte-Ladbeck, R. E., \& Meade, M. R. 1992, \aj, 103, 1652
1191: \bibitem[Whitney(1993)]{whi93} Whitney, B. A., Balm, S. P., \&
1192: Clayton, G. C. 1993, in ASP Conf. Ser. Vol. 45,
1193: Luminous High-Latitude Stars, ed. D. Sasselov (San Francisco: ASP), 105
1194: \bibitem[Whittet et al.(1992)]{whi92b} Whittet, D. C. B., Martin, P. G.,
1195: Hough, J. H., Rouse, M. F., Bailey, J. A., \& Axon, D. J. 1992,
1196: \apj, 386, 562
1197: \bibitem[Williams(1997)]{wil97} Williams, P. M. 1997, \apss, 251, 321
1198: \bibitem[Woitke, Goeres, \& Sedlmayr(1996)]{woi96} Woitke, P, Goeres, A.,
1199: \& Sedlmayr, E. 1996, \aap, 313, 217
1200: \bibitem[Wolff, Nordsieck, \& Nook(1996)]{wol96} Wolff, M. J.,
1201: Nordsieck, K. H., \& Nook, M. A. 1996, \aj, 111, 856
1202: \bibitem[Yudin et al.(2002)]{yud02} Yudin, B. F., Fernie, J. D.,
1203: Ikhsanov, N. R., Shenavrin, V. I., \& Weigelt, G. 2002, \aap, 394, 617
1204: \bibitem[Yudin et al.(2003)]{yud03} Yudin, R. V., Evans, A.,
1205: Barrett, P., Albinson, J. S., Davies, J. K., \& Hutchinson, M. G.
1206: 2003, \aap, 412, 405
1207: \bibitem[Zubko(1997)]{zub97} Zubko, V. G. 1997, \mnras, 289, 305
1208: \end{thebibliography}
1209:
1210:
1211: \clearpage
1212:
1213: \begin{figure}
1214: \plotone{f1.eps}
1215: \caption{Time variation of polarimetric properties of R CrB.
1216: This figure shows (a) visual magnitude from VSNET,
1217: (b) observed polarization amplitude,
1218: (c) its position angle, and (d) polarized flux at synthetic $V$ band
1219: against Julian date.
1220: The visual magnitude data are from VSNET database.
1221: Polarization data are shown by filled squares for observations
1222: at maximum brightness stage ($\Delta m_{\rm V} \leq 0.3$)
1223: and by open squares during declines
1224: ($\Delta m_{\rm V} > 0.3$) , with observational error ($1\sigma$).
1225: The horizontal lines in (b) and (c) denote the constant component of
1226: polarization at $V$ band (see \S 3.2).
1227: The arrow in (b) indicates that there are data points
1228: much larger than the shown range.
1229: The polarized flux is derived from variable component of
1230: polarization and magnitude drop (see \S 3.3).
1231: In the maximum brightness stage R CrB generally showed a
1232: nearly constant polarization, while it showed significant
1233: temporal variation in 2001 March and August.
1234: }
1235: \label{fig1}
1236: \end{figure}
1237:
1238: %\clearpage
1239:
1240: \begin{figure}
1241: \plotone{f2.eps}
1242: \caption{Time variation of visual magnitude and
1243: observed polarization in the epochs including
1244: the transient polarization events.
1245: For the lightcurve we show smoothed (17 point running-mean)
1246: AAVSO validated magnitudes, which includes a wealth of data
1247: points (about four times more than VSNET ones).
1248: Triangles, squares, inverse triangles, and circles denote
1249: the amplitude of observed polarization in synthetic $B$, $V$, $R$,
1250: and $I$ bands, respectively.
1251: It seems that there is temporal photometric variations
1252: of $\sim$0.1 mag a few days prior to the polarization peaks,
1253: which could be connected with small dust formation episodes.
1254: }
1255: \label{fig2}
1256: \end{figure}
1257:
1258: \begin{figure}
1259: \plotone{f3.eps}
1260: \caption{Same as Figure 2 during 2000-2001 and 2003 declines.
1261: The data are plotted with the same manner in Figure 2.
1262: The amplitude of polarization increases as the star fades.
1263: Bluer polarization increases more rapidly than red one, which
1264: is consistent with the past polarimetry.
1265: The ratio of polarization amplitude to $\Delta m_{V}$
1266: was higher in 2000--2001 decline than in 2003 decline.
1267: }
1268: \label{fig3}
1269: \end{figure}
1270:
1271: \begin{figure}
1272: \plotone{f4.eps}
1273: \caption{Sample spectra of R CrB in the maximum brightness
1274: stage including the periods of the transient polarization.
1275: The system response (instrument$+$telescope) is corrected but
1276: airmass effect is not corrected in those spectra.
1277: The date of observation is indicated at right side of each spectrum.
1278: Hatched area denotes the molecular bands (C$_{2}$).
1279: No significant variation is seen in the period.
1280: }
1281: \label{fig4}
1282: \end{figure}
1283:
1284: %\clearpage
1285:
1286: \begin{figure}
1287: \plotone{f5a.eps}
1288: \plotone{f5b.eps}
1289: \caption{Temporal variation in the wavelength dependence
1290: of the variable component of polarization at maximum
1291: brightness phases: Upper combined figure shows polarization
1292: amplitude and its position angle around the first event
1293: of the transient polarization (2001 March) and the lower one
1294: shows the second one (2001 August).
1295: The data are binned to a constant photon noise of 0.04\%.
1296: The date of observation and the $V$ magnitude below the
1297: maximum $m_{V}=6.09$ are indicated in each upper panel.
1298: %It is noted that there is little wavelength dependence
1299: %in the transient polarization.
1300: % while the $p(\lambda)$ curve
1301: %during deep declines generally has slope rising at shorter
1302: %wavelength.
1303: }
1304: \label{fig5}
1305: \end{figure}
1306:
1307: %\clearpage
1308: \begin{figure}
1309: \epsscale{.70}
1310: \plotone{f6a.eps}
1311: \plotone{f6b.eps}
1312: \plotone{f6c.eps}
1313: \caption{Same as Figure 5 during decline phases.
1314: The polarization during decline phase generally
1315: has a wavelength dependence; $p_{\rm var}$ is larger
1316: at bluer wavelengths.
1317: %in the transient polarization.
1318: % while the $p(\lambda)$ curve
1319: %during deep declines generally has slope rising at shorter
1320: %wavelength.
1321: }
1322: \label{fig6}
1323: \end{figure}
1324:
1325: %\clearpage
1326:
1327: \begin{figure}
1328: \epsscale{.70}
1329: \plotone{f7.eps}
1330: \caption{Samples of wavelength dependence of polarization
1331: calculated with Mie scattering theory.
1332: From top to bottom, we show them in the case of the scattering
1333: angle $\alpha=$45$\arcdeg$, 90$\arcdeg$ and 135$\arcdeg$,
1334: respectively.
1335: The dust model is an amorphous carbon (BE soot; \citealt{rou91}).
1336: Open squares and triangles are variable components of polarization
1337: observed on 2001 March 9 and on August 9, respectively.
1338: Dashed lines denote the results for scattering by single-sized
1339: ($a=0.07,\ 0.10,\ 0.13$ \micron ) grains.
1340: Thick solid lines are results for a mixture of grains of an
1341: inverse power-law size distribution ($n(a)\propto a^{-3.5}$)
1342: having lower and upper cutoffs at 0.005 $\micron$ and
1343: 0.25 $\micron$, respectively \citep{mat77}.
1344: In this model we assume an optically-thin case and the
1345: the dust mass corresponds to $r=2R_{\rm *}$ indicated in the
1346: parentheses should be considered as approximated values.
1347: }
1348: \label{fig7}
1349: \end{figure}
1350:
1351:
1352: %\clearpage
1353: \begin{figure}
1354: \plotone{f8.eps}
1355: \caption{Variation of {\it V} band polarization in $QU$ diagram.
1356: %Inner panel is an enlarged figure near the origin.
1357: The constant component of polarization has been subtracted.
1358: The symbols show observations at nearly maximum brightness
1359: (filled squares) and those during declines (open square).
1360: Error bar represents $1\sigma$ observational error of
1361: individual observation. The mass ejection events in 1999-2003
1362: do not seem to follow the bipolar model suggested by
1363: \citet{cla97} from past observations during decline phases,
1364: in which the data points lie at PA=0--30\arcdeg\ and 90--120\arcdeg\ on
1365: the projected sky (corresponding to PA=0--60\arcdeg\ and
1366: 180--240\arcdeg\ in this diagram).
1367: }
1368: \label{fig8}
1369: \end{figure}
1370:
1371:
1372: %\clearpage
1373: \begin{figure}
1374: \plotone{f9.eps}
1375: \caption{Geometry of the scattering model in Monte Carlo simulation.
1376: Light source ($=$ the central star) is located at the origin.
1377: The scattering cloud is a part of a spherical shell
1378: centered at the origin with an opening angle of $\phi$
1379: and having a radius of $r$.
1380: The center of the cloud is on the $x$-$z$ plane.
1381: The direction of the earth is that of $z$-axis.
1382: }
1383: \label{fig9}
1384: \end{figure}
1385:
1386:
1387: %\clearpage
1388: \begin{figure}
1389: \plotone{f10.eps}
1390: \caption{Relations between the opening angle
1391: $\phi$ and optical thickness $\tau_{0}$ satisfying $p=0.5$ \%.
1392: Thick dashed line is for $r=2 R_{*}$ and $\alpha'= 45 ^{\circ}$,
1393: thick dotted one is for $r=2 R_{*}$ and $\alpha'= 90 ^{\circ}$,
1394: thin solid one is $r=20 R_{*}$ and $\alpha'= 45 ^{\circ}$, and
1395: thin two-dashed one is $r=20 R_{*}$ and $\alpha'= 90 ^{\circ}$.
1396: The dependency of the polarization on $\tau_{0}$ become small
1397: for large $\tau_{0}$ ($\gtrsim 2$).
1398: }
1399: \label{fig10}
1400: \end{figure}
1401:
1402: %\clearpage
1403:
1404: \clearpage
1405:
1406: \begin{deluxetable}{rrrcccrcccccr}
1407: \tabletypesize{\scriptsize}
1408: \rotate
1409: \tablecolumns{13}
1410: \tablecaption{Polarimetry of R CrB. \label{tbl-1}}
1411: \tablewidth{0pt}
1412: \tablehead{
1413: \colhead{Date} & \colhead{JD} & \colhead{$\Delta m_{\rm V}$\tablenotemark{a}} &
1414: \colhead{Tel.\tablenotemark{b}} &
1415: \colhead{Diaph.\tablenotemark{c}} &
1416: \colhead{CCD\tablenotemark{d}} & \colhead{Exp.} &
1417: \colhead{$C_{2}\ 5165$\tablenotemark{e}} &
1418: \multicolumn{2}{c}{Observed polarization\tablenotemark{f}} &
1419: \multicolumn{2}{c}{Variable component\tablenotemark{f}} &
1420: \colhead{Pol. flux\tablenotemark{g}} \\
1421: \cline{9-10} \cline{11-12}
1422: \colhead{{\it yyyymmdd}} & \colhead{2450000+} &
1423: \colhead{$6.11+$} & \colhead{} &
1424: \colhead{} & \colhead{} & \colhead{sec} &
1425: \colhead{(EW) \AA} &
1426: \colhead{$p_{\rm V}$ (\%)} & \colhead{$\theta_{\rm V}$ (deg.)} &
1427: \colhead{$p_{\rm V}$ (\%)} & \colhead{$\theta_{\rm V}$ (deg.)} &
1428: \colhead{$10^{-5}F_{\rm *}$}
1429: }
1430: \startdata
1431: 19980128 & 0842.32 & $-0.06$ & D36 & D1 & T & 1200 & --- & $ 0.24\pm 0.01$ & $100.8\pm 1.8$ & $ 0.06\pm 0.02$ & $124.3\pm 6.5$& $ 64\pm 16$\\
1432: 19980325 & 0898.20 & $-0.00$ & D36 & D2 & T & 1440 & $18$ & $ 0.13\pm 0.02$ & $ 96.7\pm 4.9$ & $ 0.07\pm 0.02$ & $181.4\pm 9.1$& $ 73\pm 22$\\
1433: 19980328 & 0901.23 & $-0.03$ & D36 & D1 & T & 1680 & --- & $ 0.17\pm 0.02$ & $ 96.1\pm 3.3$ & $ 0.04\pm 0.02$ & $179.7\pm 13.4$& $ 41\pm 21$\\
1434: 19980329 & 0902.19 & $-0.08$ & D36 & D1 & T & 792 & --- & $ 0.19\pm 0.03$ & $ 96.2\pm 4.1$ & $ 0.02\pm 0.03$ & $168.9\pm 46.3$& $ 19\pm 32$\\
1435: 19980522 & 0956.07 & $-0.12$ & D36 & D1 & T & 1260 & --- & $ 0.18\pm 0.02$ & $ 98.3\pm 3.3$ & $ 0.04\pm 0.02$ & $168.0\pm 15.0$& $ 42\pm 21$\\
1436: 19980824 & 1050.01 & $+0.40$ & O74 & D1 & T & 960 & --- & $ 0.22\pm 0.01$ & $ 96.9\pm 1.3$ & $ 0.02\pm 0.01$ & $117.0\pm 11.8$& $ 16\pm 7$\\
1437: 19990222 & 1232.23 & $+1.82$ & D36 & D1 & T & 3200 & --- & $ 0.08\pm 0.02$ & $ 87.4\pm 7.4$ & $ 0.13\pm 0.02$ & $189.2\pm 4.3$& $ 25\pm 4$\\
1438: 19990223 & 1233.22 & $+1.80$ & D36 & D1 & T & 2880 & --- & $ 0.10\pm 0.02$ & $ 88.8\pm 5.3$ & $ 0.11\pm 0.02$ & $190.1\pm 4.6$& $ 21\pm 3$\\
1439: 19990225 & 1235.23 & $+1.67$ & D36 & D1 & T & 2880 & --- & $ 0.11\pm 0.01$ & $ 86.5\pm 3.5$ & $ 0.11\pm 0.01$ & $193.1\pm 3.5$& $ 23\pm 3$\\
1440: 19990227 & 1237.19 & $+1.60$ & D36 & D1 & T & 4200 & --- & $ 0.14\pm 0.01$ & $ 90.8\pm 2.9$ & $ 0.07\pm 0.01$ & $193.0\pm 5.7$& $ 16\pm 3$\\
1441: 19990316 & 1254.18 & $+0.85$ & D36 & D1 & T & 1800 & --- & $ 0.17\pm 0.01$ & $112.3\pm 1.4$ & $ 0.12\pm 0.01$ & $156.0\pm 2.1$& $ 56\pm 4$\\
1442: 19990416 & 1285.16 & $+0.38$ & D36 & D1 & T & 1800 & --- & $ 0.19\pm 0.02$ & $101.2\pm 3.3$ & $ 0.05\pm 0.02$ & $150.5\pm 14.2$& $ 32\pm 15$\\
1443: 19990501 & 1300.06 & $+0.11$ & D36 & D1 & T & 1680 & --- & $ 0.24\pm 0.01$ & $102.0\pm 1.7$ & $ 0.06\pm 0.01$ & $128.1\pm 5.8$& $ 59\pm 13$\\
1444: 19990510 & 1309.19 & $+0.09$ & O74 & D1 & T & 1440 & --- & $ 0.25\pm 0.02$ & $101.2\pm 1.6$ & $ 0.07\pm 0.01$ & $122.8\pm 6.4$& $ 61\pm 13$\\
1445: 19990517 & 1316.21 & $+0.11$ & O74 & D1 & T & 1680 & --- & $ 0.18\pm 0.02$ & $100.4\pm 3.2$ & $ 0.05\pm 0.02$ & $159.8\pm 12.1$& $ 41\pm 17$\\
1446: 19990808 & 1399.01 & $+1.09$ & D36 & D1 & S & 1500 & --- & $ 0.32\pm 0.02$ & $108.9\pm 1.8$ & $ 0.17\pm 0.02$ & $126.7\pm 3.0$& $ 60\pm 8$\\
1447: 19991202 & 1515.34 & $+1.87$ & D36 & D1 & S & 1200 & --- & $ 0.18\pm 0.04$ & $ 55.2\pm 5.9$ & $ 0.25\pm 0.04$ & $ 28.1\pm 4.4$& $ 43\pm 7$\\
1448: 19991203 & 1516.34 & $+1.87$ & D36 & D1 & S & 1200 & --- & $ 0.29\pm 0.03$ & $ 48.2\pm 4.0$ & $ 0.37\pm 0.04$ & $ 31.1\pm 3.1$& $ 64\pm 6$\\
1449: 19991207 & 1520.33 & $+2.42$ & D36 & D1 & S & 1920 & --- & $ 0.55\pm 0.02$ & $ 60.3\pm 1.2$ & $ 0.51\pm 0.02$ & $ 49.3\pm 1.3$& $ 56\pm 2$\\
1450: 20000101 & 1545.33 & $+1.16$ & D36 & D1 & S & 2400 & --- & $ 0.26\pm 0.03$ & $ 96.6\pm 3.7$ & $ 0.05\pm 0.03$ & $103.4\pm 17.2$& $ 18\pm 10$\\
1451: 20000127 & 1571.30 & $+0.72$ & D36 & D1 & S & 1560 & --- & $ 0.30\pm 0.03$ & $100.4\pm 2.3$ & $ 0.11\pm 0.03$ & $111.4\pm 6.7$& $ 55\pm 13$\\
1452: 20000203 & 1578.30 & $+0.64$ & D36 & D1 & S & 640 & --- & $ 0.33\pm 0.04$ & $ 89.4\pm 3.5$ & $ 0.14\pm 0.04$ & $ 81.1\pm 8.6$& $ 78\pm 24$\\
1453: 20000204 & 1579.26 & $+0.55$ & D36 & D1 & S & 1680 & --- & $ 0.22\pm 0.02$ & $ 97.1\pm 2.4$ & $ 0.03\pm 0.02$ & $117.4\pm 21.8$& $ 16\pm 12$\\
1454: 20000229 & 1604.19 & $+0.42$ & D36 & D1 & S & 2176 & --- & $ 0.13\pm 0.02$ & $103.1\pm 5.2$ & $ 0.09\pm 0.02$ & $171.8\pm 8.2$& $ 60\pm 14$\\
1455: 20000306 & 1610.22 & $+0.53$ & D36 & D1 & S & 2560 & --- & $ 0.18\pm 0.03$ & $105.5\pm 5.7$ & $ 0.08\pm 0.04$ & $154.1\pm 13.1$& $ 48\pm 22$\\
1456: 20000329 & 1633.15 & $+0.54$ & D36 & D1 & S & 2080 & --- & $ 0.14\pm 0.02$ & $ 78.4\pm 4.2$ & $ 0.12\pm 0.02$ & $ 25.6\pm 5.0$& $ 73\pm 13$\\
1457: 20001207 & 1886.35 & $+1.64$ & O36 & D1 & S & 1500 & --- & $ 0.64\pm 0.04$ & $ 88.4\pm 2.0$ & $ 0.44\pm 0.04$ & $ 85.5\pm 2.8$& $102\pm 10$\\
1458: 20001214 & 1893.35 & $+4.18$ & O36 & D1 & S & 2800 & --- & $ 6.81\pm 0.31$ & $102.0\pm 1.4$ & $ 6.61\pm 0.31$ & $102.2\pm 1.4$& $138\pm 6$\\
1459: 20001215 & 1894.35 & $+4.52$ & O36 & D1 & S & 2400 & --- & $ 9.98\pm 0.42$ & $ 98.4\pm 1.2$ & $ 9.78\pm 0.42$ & $ 98.5\pm 1.3$& $155\pm 7$\\
1460: 20010130 & 1940.30 & $+1.13$ & O74 & D2 & T & 5600 & $20$ & $ 0.29\pm 0.03$ & $107.0\pm 2.9$ & $ 0.13\pm 0.03$ & $127.1\pm 6.2$& $ 47\pm 11$\\
1461: 20010202 & 1943.32 & $+0.84$ & O74 & D1 & T & 2240 & --- & $ 0.15\pm 0.03$ & $132.9\pm 5.6$ & $ 0.23\pm 0.03$ & $164.2\pm 3.8$& $108\pm 14$\\
1462: 20010216 & 1957.34 & $+0.26$ & O36 & D2 & S & 1680 & $30$ & $ 0.13\pm 0.01$ & $104.9\pm 3.0$ & $ 0.10\pm 0.01$ & $171.5\pm 3.9$& $ 74\pm 11$\\
1463: 20010226 & 1967.29 & $+0.15$ & O36 & D2 & S & 2400 & $23$ & $ 0.22\pm 0.02$ & $ 85.6\pm 2.2$ & $ 0.07\pm 0.02$ & $ 51.7\pm 6.9$& $ 64\pm 16$\\
1464: 20010309 & 1978.30 & $+0.13$ & O36 & D2 & S & 1800 & $23$ & $ 0.74\pm 0.03$ & $ 92.6\pm 1.1$ & $ 0.53\pm 0.03$ & $ 91.7\pm 1.6$& $487\pm 28$\\
1465: 20010311 & 1980.17 & $+0.07$ & O36 & D1 & S & 3840 & --- & $ 0.71\pm 0.03$ & $ 85.8\pm 1.2$ & $ 0.52\pm 0.03$ & $ 82.2\pm 1.6$& $475\pm 25$\\
1466: 20010312 & 1981.24 & $+0.11$ & O36 & D2 & S & 3640 & $25$ & $ 0.50\pm 0.03$ & $ 92.1\pm 1.6$ & $ 0.30\pm 0.03$ & $ 90.2\pm 2.6$& $269\pm 26$\\
1467: 20010313 & 1982.22 & $+0.08$ & O36 & D2 & S & 2000 & $25$ & $ 0.15\pm 0.01$ & $ 98.3\pm 2.3$ & $ 0.06\pm 0.01$ & $176.6\pm 5.5$& $ 57\pm 11$\\
1468: 20010418 & 2018.09 & $+0.03$ & O36 & D2 & S & 1440 & $24$ & $ 0.13\pm 0.02$ & $ 99.6\pm 5.0$ & $ 0.08\pm 0.02$ & $177.0\pm 8.3$& $ 79\pm 23$\\
1469: 20010809 & 2131.05 & $+0.00$ & O36 & D2 & S & 900 & $22$ & $ 0.66\pm 0.06$ & $ 74.8\pm 2.5$ & $ 0.52\pm 0.06$ & $ 67.4\pm 3.1$& $520\pm 56$\\
1470: 20010811 & 2133.06 & $+0.02$ & O36 & D2 & S & 1200 & $24$ & $ 0.30\pm 0.02$ & $ 91.7\pm 2.0$ & $ 0.10\pm 0.02$ & $ 85.3\pm 5.9$& $102\pm 22$\\
1471: 20010904 & 2156.97 & $-0.04$ & O36 & D2 & S & 1080 & $28$ & $ 0.24\pm 0.02$ & $ 94.5\pm 2.5$ & $ 0.04\pm 0.02$ & $ 92.5\pm 15.6$& $ 39\pm 23$\\
1472: 20010908 & 2161.04 & $+0.05$ & O36 & D2 & S & 960 & $29$ & $ 0.21\pm 0.03$ & $ 88.4\pm 4.6$ & $ 0.05\pm 0.03$ & $ 51.3\pm 20.4$& $ 44\pm 31$\\
1473: 20010909 & 2162.03 & $+0.06$ & O36 & D2 & S & 1080 & $29$ & $ 0.12\pm 0.03$ & $ 98.4\pm 6.4$ & $ 0.09\pm 0.03$ & $179.9\pm 8.9$& $ 79\pm 26$\\
1474: 20010926 & 2178.95 & $-0.02$ & O36 & D2 & S & 1200 & $27$ & $ 0.18\pm 0.02$ & $ 94.8\pm 2.9$ & $ 0.03\pm 0.02$ & $185.4\pm 17.1$& $ 30\pm 19$\\
1475: 20011030 & 2212.90 & $-0.01$ & O36 & D2 & S & 960 & $29$ & $ 0.16\pm 0.01$ & $ 97.4\pm 2.7$ & $ 0.05\pm 0.01$ & $176.1\pm 9.1$& $ 47\pm 14$\\
1476: 20011031 & 2213.89 & $+0.00$ & O36 & D2 & S & 1120 & $29$ & $ 0.27\pm 0.02$ & $102.9\pm 2.3$ & $ 0.09\pm 0.02$ & $122.2\pm 6.7$& $ 91\pm 22$\\
1477: 20011201 & 2245.36 & $+0.12$ & O36 & D2 & S & 1600 & $22$ & $ 0.31\pm 0.01$ & $ 83.9\pm 1.2$ & $ 0.14\pm 0.01$ & $ 67.8\pm 2.6$& $131\pm 12$\\
1478: 20011202 & 2246.36 & $+0.07$ & O36 & D2 & S & 1200 & $20$ & $ 0.29\pm 0.02$ & $ 93.9\pm 2.6$ & $ 0.08\pm 0.02$ & $ 91.5\pm 9.0$& $ 75\pm 19$\\
1479: 20011227 & 2271.35 & $-0.05$ & O36 & D2 & S & 3000 & $25$ & $ 0.28\pm 0.01$ & $ 89.8\pm 1.3$ & $ 0.08\pm 0.01$ & $ 76.6\pm 4.2$& $ 81\pm 12$\\
1480: 20020122 & 2297.36 & $-0.04$ & O36 & D2 & S & 1200 & $18$ & $ 0.18\pm 0.01$ & $106.9\pm 1.6$ & $ 0.09\pm 0.01$ & $154.7\pm 3.2$& $ 85\pm 10$\\
1481: 20020124 & 2299.37 & $-0.05$ & O36 & D2 & S & 1080 & $19$ & $ 0.14\pm 0.01$ & $111.9\pm 2.4$ & $ 0.12\pm 0.01$ & $164.6\pm 2.8$& $120\pm 11$\\
1482: 20020131 & 2306.31 & $-0.03$ & O36 & D2 & S & 1800 & $25$ & $ 0.19\pm 0.01$ & $ 94.4\pm 2.0$ & $ 0.02\pm 0.01$ & $189.9\pm 22.6$& $ 17\pm 14$\\
1483: 20020206 & 2312.28 & $+0.01$ & O36 & D2 & S & 2160 & $29$ & $ 0.14\pm 0.01$ & $ 99.3\pm 2.4$ & $ 0.07\pm 0.01$ & $175.2\pm 5.1$& $ 68\pm 11$\\
1484: 20020207 & 2313.26 & $-0.01$ & O36 & D2 & S & 2160 & $28$ & $ 0.17\pm 0.01$ & $ 97.6\pm 1.6$ & $ 0.04\pm 0.01$ & $174.6\pm 5.8$& $ 44\pm 10$\\
1485: 20020224 & 2330.27 & $-0.12$ & O74 & D2 & S & 360 & $26$ & $ 0.22\pm 0.02$ & $105.1\pm 2.6$ & $ 0.08\pm 0.02$ & $140.1\pm 7.8$& $ 84\pm 21$\\
1486: 20020401 & 2366.31 & $-0.01$ & O36 & D2 & S & 1260 & $22$ & $ 0.22\pm 0.02$ & $ 98.6\pm 1.9$ & $ 0.03\pm 0.02$ & $125.0\pm 12.9$& $ 33\pm 15$\\
1487: 20020403 & 2368.22 & $-0.01$ & O36 & D2 & S & 2700 & $23$ & $ 0.27\pm 0.01$ & $ 93.8\pm 0.6$ & $ 0.07\pm 0.01$ & $ 90.6\pm 2.6$& $ 67\pm 7$\\
1488: 20020501 & 2396.11 & $+0.04$ & O36 & D2 & S & 2880 & $29$ & $ 0.22\pm 0.01$ & $ 82.4\pm 1.1$ & $ 0.09\pm 0.01$ & $ 47.6\pm 2.5$& $ 92\pm 8$\\
1489: 20020803 & 2489.97 & $-0.02$ & O36 & D2 & S & 900 & $14$ & $ 0.21\pm 0.02$ & $ 93.9\pm 2.6$ & $ 0.01\pm 0.02$ & $ 24.5\pm 51.1$& $ 10\pm 18$\\
1490: 20020804 & 2490.98 & $-0.01$ & O36 & D2 & S & 900 & $25$ & $ 0.23\pm 0.01$ & $ 90.7\pm 1.0$ & $ 0.04\pm 0.01$ & $ 67.4\pm 6.1$& $ 40\pm 9$\\
1491: 20020805 & 2492.04 & $-0.01$ & O36 & D2 & S & 900 & $24$ & $ 0.22\pm 0.02$ & $ 91.4\pm 2.1$ & $ 0.03\pm 0.02$ & $ 62.4\pm 16.6$& $ 29\pm 16$\\
1492: 20020806 & 2492.99 & $-0.00$ & O36 & D2 & S & 1260 & $25$ & $ 0.23\pm 0.01$ & $ 87.3\pm 1.5$ & $ 0.06\pm 0.01$ & $ 58.7\pm 6.1$& $ 64\pm 13$\\
1493: 20020807 & 2493.96 & $+0.01$ & O36 & D2 & S & 972 & $25$ & $ 0.23\pm 0.01$ & $ 93.1\pm 1.0$ & $ 0.03\pm 0.01$ & $ 80.6\pm 7.7$& $ 30\pm 8$\\
1494: 20030204 & 2675.33 & $+0.07$ & O36 & D2 & S & 1800 & $30$ & $ 0.23\pm 0.01$ & $ 84.4\pm 1.5$ & $ 0.08\pm 0.01$ & $ 53.8\pm 3.8$& $ 76\pm 11$\\
1495: 20030205 & 2676.31 & $+0.06$ & O36 & D2 & S & 1440 & $29$ & $ 0.22\pm 0.01$ & $ 79.7\pm 1.4$ & $ 0.11\pm 0.01$ & $ 46.4\pm 2.5$& $103\pm 10$\\
1496: 20030206 & 2677.25 & $+0.04$ & O36 & D2 & S & 1600 & $29$ & $ 0.21\pm 0.01$ & $ 84.8\pm 1.0$ & $ 0.07\pm 0.01$ & $ 45.2\pm 3.2$& $ 72\pm 7$\\
1497: 20030211 & 2682.33 & $+0.66$ & O36 & D2 & S & 3200 & $27$ & $ 0.21\pm 0.02$ & $ 97.2\pm 1.8$ & $ 0.02\pm 0.01$ & $132.8\pm 24.6$& $ 9\pm 7$\\
1498: 20030212 & 2683.29 & $+0.70$ & O36 & D2 & S & 2880 & $27$ & $ 0.20\pm 0.01$ & $101.1\pm 1.6$ & $ 0.05\pm 0.01$ & $148.0\pm 6.0$& $ 24\pm 6$\\
1499: 20030213 & 2684.32 & $+0.76$ & O36 & D2 & S & 2400 & $25$ & $ 0.30\pm 0.01$ & $103.7\pm 0.9$ & $ 0.12\pm 0.01$ & $118.8\pm 2.2$& $ 59\pm 5$\\
1500: 20030214 & 2685.31 & $+0.95$ & O36 & D2 & S & 4000 & $24$ & $ 0.30\pm 0.01$ & $104.4\pm 0.8$ & $ 0.12\pm 0.01$ & $121.0\pm 1.9$& $ 54\pm 3$\\
1501: 20030216 & 2687.34 & $+1.48$ & O36 & D2 & S & 2400 & $23$ & $ 0.34\pm 0.01$ & $120.4\pm 0.8$ & $ 0.27\pm 0.01$ & $138.9\pm 1.1$& $ 67\pm 3$\\
1502: 20030217 & 2688.30 & $+1.69$ & O36 & D2 & S & 2400 & $22$ & $ 0.42\pm 0.04$ & $126.5\pm 2.6$ & $ 0.38\pm 0.04$ & $141.0\pm 2.8$& $ 79\pm 9$\\
1503: 20030218 & 2689.34 & $+2.08$ & O36 & D2 & S & 2400 & $24$ & $ 0.67\pm 0.02$ & $127.6\pm 0.8$ & $ 0.62\pm 0.02$ & $136.4\pm 0.9$& $ 89\pm 3$\\
1504: 20030220 & 2691.28 & $+3.09$ & O36 & D2 & S & 3200 & $32$ & $ 0.97\pm 0.02$ & $141.2\pm 0.6$ & $ 1.00\pm 0.02$ & $147.1\pm 0.5$& $ 57\pm 1$\\
1505: 20030319 & 2718.26 & $+5.44$ & O36 & D2 & S & 9600 & $10$ & $ 1.64\pm 0.06$ & $ 32.5\pm 0.8$ & $ 1.77\pm 0.06$ & $ 29.7\pm 0.8$& $ 12\pm 0.4$\\
1506: 20030320 & 2719.23 & $+5.17$ & O36 & D2 & S & 10800 & $13$ & $ 1.55\pm 0.04$ & $ 35.5\pm 0.7$ & $ 1.66\pm 0.04$ & $ 32.3\pm 0.7$& $ 14\pm 0.3$\\
1507: 20030322 & 2721.19 & $+4.83$ & O36 & D2 & S & 9600 & $10$ & $ 1.23\pm 0.04$ & $ 40.6\pm 0.8$ & $ 1.31\pm 0.04$ & $ 36.3\pm 0.7$& $ 16\pm 0.5$\\
1508: 20030522 & 2782.12 & $+0.02$ & O36 & D2 & T & 4140 & $24$ & $ 0.26\pm 0.02$ & $105.7\pm 2.6$ & $ 0.10\pm 0.02$ & $129.3\pm 5.9$& $103\pm 24$\\
1509: 20030523 & 2783.02 & $+0.03$ & O36 & D2 & T & 3200 & $24$ & $ 0.27\pm 0.01$ & $110.8\pm 1.5$ & $ 0.14\pm 0.02$ & $135.9\pm 2.6$& $141\pm 15$\\
1510: 20030725 & 2845.97 & $+0.04$ & O36 & D2 & S & 700 & $44$ & $ 0.24\pm 0.01$ & $ 95.7\pm 1.6$ & $ 0.04\pm 0.01$ & $100.4\pm 10.4$& $ 38\pm 12$\\
1511: 20030727 & 2848.09 & $+0.05$ & O36 & D2 & S & 552 & $41$ & $ 0.21\pm 0.01$ & $ 94.9\pm 2.1$ & $ 0.00\pm 0.01$ & $164.1\pm 550.3$& $ 1\pm 124$\\
1512: 20030926 & 2908.96 & $-0.07$ & O36 & D2 & S & 1800 & $18$ & $ 0.24\pm 0.01$ & $ 90.4\pm 1.5$ & $ 0.05\pm 0.01$ & $ 67.8\pm 7.6$& $ 50\pm 13$\\
1513: 20030928 & 2910.92 & $-0.05$ & O36 & D2 & S & 1080 & $20$ & $ 0.24\pm 0.02$ & $ 92.5\pm 1.8$ & $ 0.04\pm 0.02$ & $ 79.5\pm 11.3$& $ 38\pm 15$\\
1514: 20030929 & 2911.92 & $-0.04$ & O36 & D2 & S & 960 & $18$ & $ 0.24\pm 0.01$ & $ 85.6\pm 1.6$ & $ 0.08\pm 0.01$ & $ 57.0\pm 3.9$& $ 77\pm 13$\\
1515: \enddata
1516:
1517: \tablenotetext{a}{Visual magnitude averaged for
1518: all data within $\pm 2.5$ days in VSNET database.}
1519: \tablenotetext{b}{D36, O36 and O74 denote 0.91 m at Dodaira,
1520: 0.91 m at Okayama and 1.88 m at Okayama, respectively.}
1521: \tablenotetext{c}{D1: 1.4 mm$\phi$ circle, D2: 0.2 mm $\times$ 1.4 mm rectangle.}
1522: \tablenotetext{d}{T and S denote TI CCD and SITe CCD, respectively.}
1523: \tablenotetext{e}{The equivalent width of the C$_{2}\ (0-0)$ 5165 \AA\
1524: band is measured between 4970 \AA\ and 5200 \AA\ for D2 observations.}
1525: \tablenotetext{f}{Observed polarization is considered as a vectorial
1526: summation of a constant component and a variable component (see text).}
1527: \tablenotetext{g}{Polarized flux $p_{\rm var} \times 10^{-0.4 \Delta m_{\rm V}}\ F_{\rm *}$}
1528:
1529:
1530: \end{deluxetable}
1531:
1532: \clearpage
1533: \begin{table*}
1534: {\scriptsize
1535: \begin{center}
1536: \begin{tabular}{lllcccc}
1537:
1538: \hline
1539: \hline
1540: $r$ & $\alpha$ & $\phi$ &
1541: \multicolumn{4}{c}{\% polarization} \\
1542: \cline{4-7}
1543: & & &
1544: $\tau_{0}=0.3$ & $\tau_{0}=1.0$ & $\tau_{0}=3.0$ & $\tau_{0}=5.0$ \\
1545: \hline
1546: $2\ R_{\rm *}$ &
1547: $45\arcdeg$ & $20\arcdeg$ & 0.035 & 0.081 & 0.118 & 0.122\\
1548: & & $30\arcdeg$ & 0.074 & 0.175 & 0.250 & 0.259\\
1549: & & $45\arcdeg$ & 0.151 & 0.351 & 0.497 & 0.511\\
1550: & & $60\arcdeg$ & 0.232 & 0.533 & 0.742 & 0.768\\
1551: \hline
1552: &
1553: $90\arcdeg$ & $20\arcdeg$ & 0.042 & 0.100 & 0.144 & 0.148\\
1554: & & $30\arcdeg$ & 0.092 & 0.218 & 0.311 & 0.321\\
1555: & & $45\arcdeg$ & 0.191 & 0.453 & 0.640 & 0.663\\
1556: & & $60\arcdeg$ & 0.306 & 0.718 & 1.014 & 1.051\\
1557: \hline
1558: &
1559: $135\arcdeg$ & $20\arcdeg$ & 0.011 & 0.026 & 0.038 & 0.039\\
1560: & & $30\arcdeg$ & 0.024 & 0.058 & 0.083 & 0.087\\
1561: & & $45\arcdeg$ & 0.052 & 0.122 & 0.174 & 0.180\\
1562: & & $60\arcdeg$ & 0.086 & 0.200 & 0.287 & 0.297\\
1563: \hline
1564: \hline
1565: $20\ R_{\rm *}$ &
1566: $45\arcdeg$ & $10\arcdeg$ & 0.030 & 0.073 & 0.110 & 0.115\\
1567: & & $20\arcdeg$ & 0.116 & 0.278 & 0.415 & 0.434\\
1568: & & $30\arcdeg$ & 0.243 & 0.579 & 0.856 & 0.892\\
1569: & & $45\arcdeg$ & 0.471 & 1.089 & 1.588 & 1.634\\
1570: \hline
1571: &
1572: $90\arcdeg$ & $10\arcdeg$ & 0.028 & 0.069 & 0.104 & 0.109\\
1573: & & $20\arcdeg$ & 0.111 & 0.269 & 0.404 & 0.422\\
1574: & & $30\arcdeg$ & 0.239 & 0.577 & 0.865 & 0.901\\
1575: & & $45\arcdeg$ & 0.489 & 1.169 & 1.732 & 1.804\\
1576: \hline
1577: &
1578: $135\arcdeg$ & $10\arcdeg$ & 0.005 & 0.012 & 0.018 & 0.019\\
1579: & & $20\arcdeg$ & 0.020 & 0.049 & 0.073 & 0.077\\
1580: & & $30\arcdeg$ & 0.045 & 0.109 & 0.164 & 0.171\\
1581: & & $45\arcdeg$ & 0.100 & 0.242 & 0.361 & 0.376\\
1582: \hline
1583: \end{tabular}
1584: \caption{Monte Carlo simulation of polarization
1585: due to scattering by dust puff\label{tbl-2}}
1586: \end{center}
1587: }
1588: \end{table*}
1589:
1590: %\clearpage
1591:
1592: \end{document}
1593:
1594: