1: %% The command below calls the preprint style
2: %% which will produce a one-column, single-spaced document.
3: %% Examples of commands for other substyles follow. Use
4: %% whichever is most appropriate for your purposes.
5:
6: %\documentclass[12pt,preprint]{aastex}
7:
8: %% manuscript produces a one-column, double-spaced document:
9:
10: % \documentclass[manuscript]{aastex}
11:
12: %% preprint2 produces a double-column, single-spaced document:
13:
14: \documentclass[preprint2]{aastex}
15:
16:
17: \begin{document}
18:
19: \title{Butterfly Diagram and Activity Cycles in HR 1099}
20:
21: \author{Svetlana V. Berdyugina}
22: \affil{Institute of Astronomy, ETH Zurich, 8092 Zurich, Switzerland}
23: \affil{Tuorla Observatory, University of Turku, V\"ais\"al\"antie 20, FIN-21500, Piikki\"o, Finland}
24: \email{sveta@astro.phys.ethz.ch}
25:
26: \and
27: \author{Gregory W. Henry}
28: \affil{Center of Excellence in Information Systems,
29: Tennessee State University, Nashville, TN 37203}
30: \email{henry@schwab.tsuniv.edu}
31:
32: \begin{abstract}
33: We analyze photometric data of the active RS CVn--type star HR~1099 for the
34: years 1975--2006 with an inversion technique and reveal the nature of
35: two activity cycles of 15--16 yr and 5.3$\pm$0.1 yr duration. The 16 yr
36: cycle is related to variations of the total spot area and is coupled with
37: the differential rotation, while the 5.3 yr cycle is caused by the symmetric
38: redistribution of the spotted area between the opposite stellar hemispheres
39: (flip-flop cycle). We recover long-lived active regions comprising two
40: active longitudes that migrate in the orbital reference frame with a variable
41: rate because of the differential rotation along with changes in the mean
42: spot latitudes. The migration pattern is periodic with the 16 yr cycle.
43: Combining the longitudinal migration of the active regions with a
44: previously measured differential rotation law, we recover the first stellar
45: butterfly diagram without an assumption about spot shapes.
46: We find that mean latitudes of active regions at opposite longitudes change
47: antisymmetrically in the course of the 16 yr cycle: while one active region
48: migrates to the pole, the other approaches the equator. This suggests a
49: precession of the global magnetic field with respect to the stellar rotational
50: axis.
51: \end{abstract}
52:
53: \keywords{stars: activity, imaging, rotation, spots,
54: individual (HR~1099)---binaries: spectroscopic---techniques: photometric}
55:
56:
57: %***********************************************************************
58: \section{Introduction}
59: %***********************************************************************
60:
61: HR~1099 (= V711~Tau) is the most extensively studied RS~CVn system,
62: consisting
63: of a K1 IV primary and a G5 IV secondary in a 2.8 day orbit. The primary,
64: being the faster rotator, is the first cool active star whose surface was
65: mapped with the Doppler imaging (DI) technique \citep{vogt&pen83}. It has
66: a record 46 surface maps obtained between 1981 and the present
67: \citep{don92,don03,don99,vogt99,str&bar00,gar03,pet04}. Almost all images
68: of the star feature a prominent, cool, polar cap along with a variety of
69: high- and low-latitude spots. The area of the polar spot and its extensions
70: may vary cyclically and correlate with the mean brightness of the system
71: \citep{vogt99}.
72:
73: The photometric variability of HR~1099 was discovered in 1975 by
74: \citet{lan&hal76} and has been followed regularly for more than 30 years.
75: The data before 1993 were collected and analyzed by \citet{hehh95}. They
76: found periodicities of 5.5$\pm$0.3 and 16$\pm$1 yr, respectively,
77: in the star's mean
78: and maximum brightness, suggesting the presence of activity cycles. Using
79: a model with two circular spots, they identified 15 different dark spots
80: within the 18 yr span of the data with lifetimes averaging between 1.5
81: and 2.0 yr. The spot rotation periods ranged from 2.8245 to 2.8412 days
82: and bracketed the orbital period by $-$0.47\% to +0.12\% days, implying
83: significant differential rotation. \citet{vogt99} measured the differential
84: rotation from DI and found that it is of opposite sign and about a factor of
85: 56 less than for the Sun. With newer stellar images, however, \citet{pet04}
86: argued that the differential rotation law for HR~1099 is of the same sign
87: as the Sun and a factor of 2 larger than determined by \citet{vogt99}.
88: The range of polar and equatorial periods determined by \citet{pet04}
89: encompasses all periods derived from the photometry of \citet{hehh95}.
90:
91: \citet{moh&rav93} interpreted the photometric variability of HR~1099 as
92: the result of two active longitude regions migrating in the orbital reference
93: frame with constant but very different rates. Two active longitudes (ALs)
94: separated by about 180\degr\ and locked in synchronism with the orbital
95: period have been surmised by \citet{hehh95}. In comparison with other
96: RS~CVn--type stars, however, the ALs on HR~1099 appeared very broad,
97: 110\degr--140\degr, implying smearing by differential rotation.
98:
99: In the present Letter we analyze all photometric data collected up to date
100: for HR~1099, previously published and newly obtained. Employing a technique
101: based on inversions of stellar light curves into images of the spot filling
102: factor \citep{ber02}, we recover locations of active regions and search for
103: long-lived ALs. Taking into account the effect of differential rotation,
104: we are able to recover a stellar butterfly diagram, which, in contrast to
105: previous studies \citep[e.g.,][]{kat03}, does not involve assumptions about
106: spot distributions, shapes, and numbers.
107:
108: %***********************************************************************
109: \section{Observations}\label{sec:obs}
110: %***********************************************************************
111:
112: We have analyzed the Johnson $V$ photometric observations collected earlier
113: from the literature and from the T1 0.25~m and T3 0.40~m automatic photometric
114: telescopes (APTs) by \citet{hehh95}, that span the years 1975--1993. We
115: have also included new observations for the years 1994--2006, acquired with
116: the T3 APT. The T1 and T3 APTs, along with several others, are housed at
117: Fairborn Observatory in southern Arizona. Further details on the operation
118: of the APTs and the reduction of the data can be found in \citet{h95a,h95b}
119: and \citet{ehf03}. The complete set of Johnson $V$ differential magnitudes
120: is listed in Table~\ref{tab:phot} and plotted in Fig.~\ref{fig:alon}. The
121: differential magnitudes have been converted to apparent magnitudes in
122: Fig.~\ref{fig:alon}$a$ with a $V$ magnitude of 4.28 for the comparison star
123: 10~Tau (= HR~1101). The faint, close, visual companion of HR~1099 was always
124: included in the photometer diaphragm and so is included in the light of
125: HR~1099.
126:
127: The $V$ observations were divided into 109 subsets
128: within which the light curve was observed to be stable over each
129: time interval. The average length of each subinterval was $\leq$40 days.
130: Julian Dates of the observations were converted to orbital phases with the
131: ephemeris
132: \begin{equation}
133: T_{\rm conj}=2,442,766.08+2.83774E
134: \label{eph}
135: \end{equation}
136: adopted from \citet{fek83}, where zero phase corresponds to the conjunction
137: with the K1 star closest to the Earth.
138: The error in the period of 10$^{-5}$\,days results in a phase uncertainty
139: of only 0.04 after 30 years.
140: According to the Fourier analysis of \citet{hehh95}, the star exhibits the
141: presence of a small ellipticity effect with a full amplitude of 0.016 mag in
142: $V$. This effect was removed from each data set before the light-curve
143: inversions were performed.
144:
145: %***********************************************************************
146: \section{Activity cycles and active longitudes}\label{sec:cyc}
147: %***********************************************************************
148:
149: For each of the 109 light curves, we measure the maximum, mean, and minimum
150: $V$ magnitudes as well as the peak-to-peak amplitude in $V$. Period analysis
151: of these values confirms the cycles found by \citet{hehh95}. The mean
152: brightness variations show a cycle of $\sim$15 years (Fig.~\ref{fig:alon}$a$),
153: and similar cycle lengths of 15--16 years are seen in the maximum and minimum
154: magnitudes.
155: %The two brightest states of the star, $V=5\fm65$ and $5\fm67$, were
156: %observed in 1984 and 2000. The two faintest states, $5\fm94$ and $5\fm95$,
157: %occurred in 1978 and 1994, respectively, being separated by 15.3 years.
158: The peak-to-peak amplitude in $V$ varies with a cycle of 5.3$\pm$0.1 yr
159: (Fig.~\ref{fig:alon}$b$). This cycle also has an influence on variations
160: in the maximum, mean, and minimum $V$ magnitudes. Similarly, the
161: 16 yr cycle also influences the amplitude variations, as the largest
162: amplitude of 0.2 mag was observed at the maxima of the spottedness in
163: 1979 and 1994.
164:
165: To determine the large-scale spot distribution on HR~1099, we apply an
166: inversion technique to the light curves using a two-temperature approximation
167: \citep{ber02}.
168: The inversion of the light curve results in the distribution of the spot
169: filling factor over the stellar surface, i.e.,\ a stellar image
170: (Fig.~\ref{fig:image}).
171: We adopted values of 4700 and 3500\,K for the photosphere and spot
172: temperatures, respectively, following \citet{hat96}. Also, we assumed
173: that the star is spotless at $V=5.6$ mag, which is about 0.05 mag brighter
174: than the brightest magnitude observed for HR~1099. The inclination of the
175: rotational axis to the line of sight is chosen to be 40$^\circ$;
176: inclinations in the range of 33\degr--45\degr\ have been used for
177: the previous DI mentioned above.
178:
179: From the stellar images, we recover spot longitudes with the maximum spot
180: filling factor
181: \citep[see details in][]{ber02}. The accuracy of the spot longitudes
182: depends on the broadness of the light curve minimum and is on average 0.05
183: in phase.
184: When two minima are seen in the light curves, two spot concentrations are
185: recovered in the images. The time dependence of the spot phases is shown in
186: Fig.~\ref{fig:alon}$c$ (the data are available on request). Our derived
187: spot phases constitute two sequences that represent two migrating active
188: longitude regions separated on average by 180$^\circ$, similar to
189: the previous studies of HR~1099 \citep{moh&rav93,hehh95,lan06} and to
190: other RS~CVn--type stars \citep{ber&tuo98,fek&hen05}.
191: \cite{lan06} related the long-term trend in the migration of the ALs
192: with binary orbital period variations.
193:
194: We notice that the two active regions regularly approach each other and
195: merge into a large single region. Subsequently, the single region divides
196: back into two spot regions. Thus, we are seeing the crossing of the migration
197: paths of the two ALs.
198: In the presence of differential rotation, such behavior suggests
199: variations in mean spot latitudes.
200: Three intersections were clearly observed in 1979, 1990, and 1995. At the
201: same times, the amplitude of the brightness variations reached a maximum
202: (Fig.~\ref{fig:alon}$b$), which implies a rearrangement of spots onto one
203: stellar hemisphere and supports the idea of merging active regions and
204: crossing ALs.
205: The intersections in 1979 and 1990 are also seen in the reconstruction
206: by \cite{lan06}. However, they missed the one in 1995 due to lack of data.
207: The shorter ($\sim$5 yr) and longer ($\sim$11 yr) time intervals between
208: the intersections comprise the 16 yr cycle.
209:
210: In Fig.~\ref{fig:alon}$c$, the two migration paths are plotted with
211: two fits to
212: the observed phases. The fits are obtained with a two-harmonic Fourier
213: series and a long-term trend, which together are able to reproduce the
214: nonsymmetric shapes of the curves.
215: The assignment of the active regions to one or the other AL has been
216: done under the assumption that the active region continues its migration
217: with the same rate after the intersection of the ALs, which is natural in
218: the presence of differential rotation.
219: Although the fits have been calculated separately for the two ALs, both
220: reveal similar shapes that are mirrored with respect to each other in
221: phase. The migration patterns appear to be periodic with a 16$\pm$3
222: yr cycle.
223: The standard deviation of the fits is as small as 0.06 phase units. An
224: extrapolation of the fits suggests that the next crossing of the active
225: longitudes will occur in 2007. Observation of the ALs crossing in 2007
226: with an increase in photometric amplitude will be a good test for the
227: present model.
228:
229: In addition to the migration patterns and their intersections, abrupt
230: switches of the dominant activity between the ALs occurred in 1988,
231: 1992--1993,
232: 1998--1999, and 2005, all at the times when the amplitude of the light curve
233: was near the minimum. Such events periodically occur in RS~CVn-- and
234: FK~Com--type stars and are known as flip-flop cycles \citep{ber&tuo98,kor02}.
235: In HR~1099, the flip-flops occurred when the active regions at opposite
236: longitudes became about the same size; that reduced the amplitude of the
237: light curve, and the dominate activity switched to the opposite longitude.
238: Thus, flip-flops in HR~1099 are apparently governed by the 5.3 yr cycle but
239: appear to be disturbed by the differential rotation and intersections of the
240: active longitudes.
241:
242: %***********************************************************************
243: \section{The butterfly diagram}\label{sec:fly}
244: %***********************************************************************
245:
246: The shape of the spot migration curves for HR~1099 is reminiscent of the
247: solar ALs \citep{ber&us03}. On the Sun, such a migration pattern is
248: determined
249: by changes of the mean spot latitudes along with the surface differential
250: rotation. Assuming that a similar relation holds for HR~1099, we can recover
251: mean spot latitudes on its surface, i.e., create butterfly diagrams for the
252: active longitudes in the stellar hemisphere visible above the stellar
253: equator. For this purpose we use the differential rotation law:
254: \begin{equation}
255: \Omega=\Omega_{\rm eq} - \Delta\Omega\sin^2\psi,
256: \label{eq:dif}
257: \end{equation}
258: with the parameters estimated by \citet{pet04} from DI of HR~1099 for
259: the year 2002, having the smallest variances:
260: $\Omega_{\rm eq}$=2.2241$\pm$0.0004\,rad\,day$^{-1}$ and
261: $\Delta\Omega$=15.2$\pm$0.8\,mrad\,day$^{-1}$.
262: By observing a phase shift of the ALs during a given time interval, we can
263: determine an instant angular velocity $\Omega$, and, using Eq.~(\ref{eq:dif}),
264: we can find the corresponding latitude $\psi$. To increase accuracy, we
265: determine phase shifts from the harmonic fits to the migration paths. The
266: results are shown as solid and dashed lines in
267: Fig.~\ref{fig:butterfly}$a$~\&~$b$. As expected, mean spot latitudes
268: vary with the 16 yr cycle for both active longitudes.
269:
270: The extent and shape of the butterfly diagram recovered from the AL
271: migration can be compared with the measurements of spot latitudes from DI.
272: From the 46 published images, we measured latitude extents of the recovered
273: spots and calculated weighted mean latitudes for each of the two opposite
274: ALs. These are shown in Fig.~\ref{fig:butterfly}$a$~\&~$b$. The mean DI
275: latitudes and the curves obtained from the AL migration are in good
276: agreement, thus supporting our interpretation of the photometry.
277: To estimate a possible range of the parameters in Eq.~(\ref{eq:dif}),
278: we searched for the best fit (via $\chi^2$ minimization) to the mean DI
279: latitudes combined with the angular velocity $\Omega$ measured from the
280: AL migration. The result is shown in Fig.~\ref{fig:difrot}. It reveals a
281: very elongated error ellipsoid encompassing measurements by \citet{pet04}
282: for different years.
283: Our best-fit parameters, $\Omega_{\rm eq}$=2.230$\pm$0.003\,rad\,day$^{-1}$
284: and $\Delta\Omega$=22$\pm$4\,mrad\,day$^{-1}$, indicate a somewhat stronger
285: differential rotation on HR~1099 than the best parameters by \citet{pet04},
286: although the two measurements appear to be statistically equivalent.
287:
288: The shape of the butterfly diagram for HR~1099 is reminiscent of the Sun's.
289: The direction of the latitude migration, however, is different for the two
290: ALs. For the active region shown in Fig.~\ref{fig:alon}c with a solid line,
291: preferred spot appearances first move relatively rapidly (for 5 years) to
292: higher latitudes, up to 70$^\circ$, and then slowly (for 11 years) move back
293: to mid latitudes, down to 40$^\circ$ (Fig.~\ref{fig:butterfly}$a$). For the
294: other active region (shown in Fig.~\ref{fig:alon}$c$ with a dashed line),
295: the behavior of the spot latitudes is just the opposite.
296: %During the first 5 years of the cycle, preferred spot appearances seem to
297: %move to lower latitudes, down to 40$^\circ$, and then for the next 11 years,
298: %they move to higher latitudes, up to 70$^\circ$ (Fig.~\ref{fig:butterfly}b).
299: The latitude shifts are therefore always opposite for the two ALs.
300: During 1982--1987, both ALs were at about the same latitude, and the
301: migration rate remained constant. In 1988--1993, one active longitude was
302: occupied by large high-latitude (polar) spots, while low-latitude spots
303: appeared around the opposite active longitude. After 1997, the dominant
304: active regions again moved to about the same latitude, thus completing the
305: 16 yr cycle. The opposite latitudinal migration of the two long-lived
306: regions could be responsible for the contradictory measurements of the
307: differential rotation by \citet{vogt99} and \citet{pet04}.
308:
309: %***********************************************************************
310: \section{Discussion}\label{sec:dis}
311: %***********************************************************************
312:
313: As noted by \citet{hehh95}, if the 16 yr cycle is akin to the 11 yr sunspot
314: cycle, one might anticipate other signatures of such a cycle. Indeed, the
315: 16 and 5.3 yr activity cycles in HR~1099 are similar to the sunspot cycles
316: of 11 and 3.7 yr \citep{ber&us03} and the cycles observed on young solar
317: analogs \citep{ber&jar05}. The longer cycle in HR~1099 reflects the variation
318: of the total spottedness and is coupled with the differential rotation.
319: This cycle may be associated with an oscillating axisymmetric dynamo mode
320: in which global polarity reversals occur in cycles of 22 years on the Sun or
321: 32 years on HR~1099. No reversals have yet been observed on HR~1099 despite
322: 12 years of spectropolarimetric observations \citep{don03}. Comparing the
323: butterfly diagrams of the Sun and HR~1099, we suppose that a global polarity
324: reversal on HR~1099 could have happened around 1990, just before the regular
325: spectropolarimetric observations by Donati et al.\ began. The next one can
326: be expected in 2007.
327:
328: %The long-term trend in the migration of active regions is probably
329: %related to the orbital period variations detected in the HR~1099 binary
330: %\citep{don99,lan06}. Before the year 1987 the orbital period was nearly
331: %constant, while the active regions were migrating to later orbital phases.
332: %After 1987, the period was steadily decreasing but the AL
333: %stayed on average at the same phases.
334: %This was also discussed by \cite{lan06}.
335: %This gives
336: %further support to the possible connection between the stellar gravitational
337: %quadrupole moment, which varies with magnetic activity, and orbital
338: %period modulation, as proposed by \citet{app92}.
339:
340: The shorter activity cycle on HR~1099 is related to the flip-flop phenomenon;
341: i.e., it reflects symmetrical redistribution of the spotted area between
342: opposite stellar hemispheres. In the light curves, the flip-flops are seen
343: as a strengthening of the secondary minima and a diminishing of the primary
344: one. In the Doppler images of HR~1099 obtained by \citet{don99} and
345: \citet{don03}, the flip-flop in 1992--1993 is seen as an abrupt change of the
346: phase towards which the polar spot is tilted by about half a rotational
347: period, while the flip-flop in 1998--1999 appears as a progressive shift of
348: the polar spot extension to the opposite longitude. The flip-flop cycles
349: can be associated with a non-axisymmetric dynamo mode, which gives rise to
350: ALs \citep{fluri&berd04}. Similar to the Sun and young solar analogs, the
351: flip-flop cycle on HR~1099 is about one-third of the length of the cycle
352: regulating the total spottedness, which implies a certain relation
353: between the non- and axisymmetric dynamo modes.
354:
355: The fact that the butterfly diagrams for the two ALs are antisymmetric, i.e.,
356: that latitude shifts are always opposite in the opposite stellar hemispheres,
357: implies that the axis perpendicular to the line connecting the ALs changes
358: its direction with respect to the rotational axis by about $\pm$15\degr\
359: during the 16 yr cycle. Thus, the two axes are separated the most in
360: 1990--1991 and 1995--1996 and aligned in 1982--1987 and 1998--2003,
361: suggesting the
362: precession of the nonaxisymmetric dynamo mode during the 16 yr cycle.
363: Such a precession of the magnetic field was also suspected in the
364: Zeeman-Doppler images of HR~1099 obtained by \citet{don03}, who noticed
365: that the rings of the azimuthal field got progressively better centered in
366: 2000--2002, while the fraction of the magnetic energy attributed to the
367: meridional field component was regularly increasing in 1992--1997 and
368: decreasing starting from 1998 up to now. They assumed that such behavior
369: suggests that the main axis of symmetry of the azimuthal field component is
370: varying in orientation with time on the timescale of at least one decade,
371: which is in agreement with our findings.
372:
373: Our results suggest that a few interesting events in the magnetic activity of
374: HR~1099 may happen in 2007. For instance, an extrapolation of the fits to
375: the ALs predicts their next crossing to happen during this time. Also, we
376: can expect the mean spot latitude of the dominant active region to be
377: different again and the rings of the azimuthal field to be decentered. In
378: addition, a global polarity reversal may finally be observed. The detection
379: of such events would be of great importance for understanding stellar magnetic
380: cycles and their underlying dynamo processes.
381:
382: We thank Lou Boyd at Fairborn for making our automated photometry possible.
383: This research at Tennessee State University was supported by NASA grant
384: NCC5-511 and NSF grant HRD-9706268. SVB acknowledges the EURYI (European Young
385: Investigator) Award provided by the European Science Foundation
386: (see www.esf.org/euryi).
387:
388: \begin{thebibliography}{}
389:
390: %\bibitem[Applegate(1992)]{app92}
391: %Applegate, J. H. 1992, \apj, 385, 621
392:
393: %\bibitem[Berdyugina(1998)]{ber98}
394: %Berdyugina, S. V. 1998, \aap, 338, 97
395:
396: \bibitem[Berdyugina \& J\"arvinen(2005)]{ber&jar05}
397: Berdyugina, S. V., \& J\"arvinen, S. P. 2005, Astron. Nachr., 326, 283
398:
399: \bibitem[Berdyugina et al.(2002)]{ber02}
400: Berdyugina, S. V., Pelt, J., \& Tuominen, I. 2002, \aap, 394, 505
401:
402: \bibitem[Berdyugina \& Tuominen(1998)]{ber&tuo98}
403: Berdyugina, S. V., \& Tuominen, I. 1998, \aap, 336, L25
404:
405: \bibitem[Berdyugina \& Usoskin(2003)]{ber&us03}
406: Berdyugina, S. V., \& Usoskin, I. G. 2003, \aap, 405, 1121
407:
408: \bibitem[Donati(1999)]{don99}
409: Donati, J.-F. 1999, \mnras, 302, 457
410:
411: \bibitem[Donati et al.(1992)]{don92}
412: Donati, J.-F., Brown, S. F., Semel, M., Rees, D. E., Dempsey, R. C., Matthew,
413: J. M., Henry, G. W., \& Hall, D. S. 1992, \aap, 265, 682
414:
415: \bibitem[Donati et al.(2003)]{don03}
416: Donati, J.-F., et al. 2003, \mnras, 345, 1145
417:
418: \bibitem[Eaton et al.(2003)]{ehf03}
419: Eaton, J. A., Henry, G. W., \& Fekel, F. C. 2003, in The Future of
420: Small Telescopes in the New Millennium, Volume II - The Telescopes
421: We Use, ed. T. D. Oswalt (Dordrecht: Kluwer), 189
422:
423: \bibitem[Fekel(1983)]{fek83}
424: Fekel, F. C. 1983, \apj, 268, 274
425:
426: \bibitem[Fekel \& Henry(2005)]{fek&hen05}
427: Fekel, F. C., \& Henry, G. W. 2005, \aj, 129, 1669
428:
429: \bibitem[Fluri \& Berdyugina(2004)]{fluri&berd04}
430: Fluri, D. M., \& Berdyugina, S. V. 2004, \solphys, 224, 153
431:
432: %\bibitem[Frasca \& Lanza(2005)]{fra&lan05}
433: %Frasca, A., \& Lanza, A. F. 2005, \aap, 429, 309
434:
435: \bibitem[Garc\'{\i}a-Alvarez et al.(2003)]{gar03}
436: Garc\'{\i}a-Alvarez, D., Barnes, J. R., Collier Cameron, A., Doyle, J. G.,
437: Messina, S., Lanza, A. F., \& Rodon\`o, M. 2003, \aap, 402, 1073
438:
439: \bibitem[Hatzes et al.(1996)]{hat96}
440: Hatzes, A. P., Vogt, S. S., Ramseyer, T. F., \& Misch, A. 1996, \apj, 469, 808
441:
442: \bibitem[Henry(1995a)]{h95a}
443: Henry, G. W. 1995a, in ASP Conf. Ser. 79, Robotic Telescopes: Current
444: Capabilities, Present Developments, and Future Prospects for Automated
445: Astronomy, ed. G. W. Henry \& J. A. Eaton (San Francisco: ASP), 37
446:
447: \bibitem[Henry(1995b)]{h95b}
448: Henry, G. W. 1995b, in ASP Conf. Ser. 79, Robotic Telescopes: Current
449: Capabilities, Present Developments, and Future Prospects for Automated
450: Astronomy, ed. G. W. Henry \& J. A. Eaton (San Francisco: ASP), 44
451:
452: \bibitem[Henry et al.(1995)]{hehh95}
453: Henry, G. W., Eaton, J. A., Hamer, J., \& Hall, D. S. 1995, \apj, 97, 513
454:
455: %\bibitem[J\"arvinen et al.(2005a)]{jar05a}
456: %J\"arvinen, S. P., Berdyugina S. V., Tuominen, I., Cutispoto, G., \& Bos, M.
457: % 2005a, \aap, 432, 657
458:
459: %\bibitem[J\"arvinen et al.(2005b)]{jar05b}
460: %J\"arvinen, S. P., Berdyugina S. V., \& Strassmeier, K. G. 2005b, \aap,
461: % 440, 735
462:
463: \bibitem[Katsova et al.(2003)]{kat03}
464: Katsova, M. M., Livshits, M. A., \& Belvedere, G. 2003, Sol. Phys., 216, 353
465:
466: \bibitem[Korhonen et al.(2002)]{kor02}
467: Korhonen, H., Berdyugina, S. V., \& Tuominen, I. 2002, \aap, 390, 179
468:
469: \bibitem[Landis \& Hall(1976)] {lan&hal76}
470: Landis, H. J., \& Hall, D. S. 1976, Inf. Bull. Var. Stars, No. 1113
471:
472: \bibitem[Lanza et al.(2006)]{lan06}
473: Lanza, A. F., Piluso, N., Rodoṇ, M., Messina, S., \& Cutispoto, G. 2006,
474: \aap, 455, 595
475:
476: \bibitem[Mohin \& Raveendran(1993)] {moh&rav93}
477: Mohin, S., \& Raveendran, A. V. 1993, \aaps, 100, 331
478:
479: \bibitem[Petit et al.(2004)]{pet04}
480: Petit, P., et al. 2004, \mnras, 348, 1175
481:
482: \bibitem[Strassmeier \& Bartus(2000)]{str&bar00}
483: Strassmeier, K. G., \& Bartus, J. 2000, \aap, 354, 537
484:
485: \bibitem[Vogt \& Penrod(1983)]{vogt&pen83}
486: Vogt, S. S., \& Penrod, G. D. 1983, \pasp, 95, 565
487:
488: \bibitem[Vogt et al.(1999)]{vogt99}
489: Vogt, S. S., Hatzes, A. P., Misch, A. A., \& K\"urster, M. 1999, \apjs,
490: 121, 547
491:
492: \end{thebibliography}
493:
494:
495: %\clearpage
496:
497: \begin{figure}
498: \plotone{f1.eps}
499: \caption{Activity patterns on HR~1099.
500: ($a$) All $V$-magnitude observations analyzed in this Letter.
501: The data before 1993 are from \citet{hehh95}. Cyclic variations of the
502: mean brightness are revealed with a harmonic fit ({\emph solid curve})
503: with a period of $\sim$15 years.
504: ($b$) Peak-to-peak amplitude of the light curves
505: ({\emph filled circles}). A harmonic fit ({\emph solid curve}) reveals
506: the period of 5.3$\pm$0.1 yr.
507: ($c$) Phases of the spots recovered with the light-curve
508: inversions. Filled circles denote the larger or single active region
509: observed at a given time, while open circles indicate a second, smaller
510: region if observed. Migration paths of the ALs are
511: indicated with solid and dashed curves, which are harmonic fits to
512: the measurements with a long-term trend. The migration pattern is
513: apparently repeated in about 16 years.
514: }
515: \label{fig:alon}
516: \end{figure}
517:
518: %\clearpage
519:
520: \begin{figure}
521: \plotone{f2.eps}
522: \caption{Examples of the spot filling factor distribution on HR~1099
523: ({\emph left}) and corresponding light curves ({\emph right}).
524: The stellar surface was modeled
525: with a grid of 10$^{\circ}\times$10$^{\circ}$. The spot filling factor at
526: each grid point indicates relative contributions from two temperature
527: components, the hot photosphere and cool spots. In the images, the darker
528: area implies a higher spot filling factor. Observed light curves are shown
529: as crosses and the fit as the solid line. In contrast to direct modeling
530: of light curves, the inversions do not involve any assumptions about spot
531: shapes, numbers, or latitudes.
532: }
533: \label{fig:image}
534: \end{figure}
535:
536: %\clearpage
537:
538: \begin{figure}
539: \plotone{f3.eps}
540: \caption{Butterfly diagram of HR~1099.
541: ($a$) Mean latitudes recovered from the fit to the active longitude
542: shown in Fig.~\ref{fig:alon}$c$ as the solid curve.
543: ($b$) Mean latitudes recovered from the fit to the active longitude
544: shown in Fig.~\ref{fig:alon}$c$ as the dashed curve.
545: In both plots, the mean latitudes of spots in the opposite hemispheres
546: measured from DI are shown with different symbols
547: corresponding to the following sources:
548: VP83 \citep{vogt&pen83},
549: V99 \citep{vogt99},
550: D92 \citep{don92},
551: D99 \citep{don99},
552: SB00 \citep{str&bar00},
553: G03 \citep{gar03},
554: D03 \citep{don03},
555: P04 \citep{pet04}, and
556: P06 (P.~Petit 2006, private communication).
557: Standard deviations of the mean DI latitudes (if several spots were
558: observed) are shown as vertical dashes. The thick curves are the solutions
559: using the best parameters of the differential rotation law by \cite{pet04}.
560: The thin curves indicate the $\pm$1\,$\sigma$ interval for the best parameters
561: obtained in this Letter (filled circle in Fig.~\ref{fig:difrot}).
562: The two solutions are statistically equivalent.
563: }
564: \label{fig:butterfly}
565: \end{figure}
566:
567: \begin{figure}
568: \plotone{f4.eps}
569: \caption{The $\chi^2$ surface contours for the two parameters in
570: Eq.~(\ref{eq:dif}) calculated from the AL migration and the DI mean spot
571: latitudes. The $\chi^2$=1 level is emphasized by a thick contour. The
572: best-fit solution is indicated by the filled circle. The measurements
573: by \cite{pet04} are shown as open circles. The strongly elongated shape
574: of the minimum implies that the two parameters are almost linearly dependent.
575: The distribution of the measurements along the major axis of the error
576: ellipsoid suggests that they reflect errors in DI spot latitudes rather
577: than true variations of the differential rotation.
578: }
579: \label{fig:difrot}
580: \end{figure}
581:
582: %\clearpage
583:
584: \begin{deluxetable}{ccc}
585: \tablewidth{0pt}
586: \tablecaption{HR~1099 Photometric Data.
587: \label{tab:phot}}
588: \tablehead{
589: \colhead{HJD} & \colhead{$\Delta$V} }
590: \startdata
591: 2,442,719.8650& 1.600\\
592: 2,442,720.9300& 1.510\\
593: 2,442,721.9050& 1.570\\
594: \enddata
595: \tablenotetext{*}{Table 1 is published in its entirety in the electronic
596: edition of the {\em Astrophysical Journal}. A portion is shown here for
597: guidance regarding its form and content.}
598: \end{deluxetable}
599:
600: \end{document}
601: