1: % mn2esample.tex
2: %
3: % v2.1 released 22nd May 2002 (G. Hutton)
4: %
5: % The mnsample.tex file has been amended to highlight
6: % the proper use of LaTeX2e code with the class file
7: % and using natbib cross-referencing. These changes
8: % do not reflect the original paper by A. V. Raveendran.
9: %
10: % Previous versions of this sample document were
11: % compatible with the LaTeX 2.09 style file mn.sty
12: % v1.2 released 5th September 1994 (M. Reed)
13: % v1.1 released 18th July 1994
14: % v1.0 released 28th January 1994
15:
16: \documentclass[useAMS,usenatbib,usegraphicx]{mn2e}%
17: %\documentclass[useAMS,usenatbib,usegraphicx,referee]{mn2e}%
18:
19: % If your system does not have the AMS fonts version 2.0 installed, then
20: % remove the useAMS option.
21: %
22: % useAMS allows you to obtain upright Greek characters.
23: % e.g. \umu, \upi etc. See the section on "Upright Greek characters" in
24: % this guide for further information.
25: %
26: % If you are using AMS 2.0 fonts, bold math letters/symbols are available
27: % at a larger range of sizes for NFSS release 1 and 2 (using \boldmath or
28: % preferably \bmath).
29: %
30: % The usenatbib command allows the use of Patrick Daly's natbib.sty for
31: % cross-referencing.
32: %
33: % If you wish to typeset the paper in Times font (if you do not have the
34: % PostScript Type 1 Computer Modern fonts you will need to do this to get
35: % smoother fonts in a PDF file) then uncomment the next line
36: % \usepackage{Times}
37:
38: %%%%% AUTHORS - PLACE YOUR OWN MACROS HERE %%%%%
39:
40:
41: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 07-06-07
42:
43: \title[Periodic flaring rate on YY Gem]{Simulations of the periodic flaring rate on YY Gem}
44: \author[D. H. Gao, P. F. Chen, M. D. Ding, and X. D. Li]
45: {D. H. Gao\thanks{E-mail: mg0526001@smail.nju.edu.cn},
46: P. F. Chen\thanks{chenpf@nju.edu.cn},
47: M. D. Ding\thanks{dmd@nju.edu.cn},
48: X. D. Li\\
49: Department of Astronomy, Nanjing University, Nanjing 210093,
50: China}
51:
52: \begin{document}
53:
54: \date{Accepted --. Received --; in original form --}
55:
56: \pagerange{\pageref{firstpage}--\pageref{lastpage}} \pubyear{2007}
57:
58: \maketitle
59:
60: \label{firstpage}
61:
62: \begin{abstract}
63: The binary YY Gem shows many interesting properties, one of which
64: is the periodicity in its flaring rate. The period, which is about
65: $48 \pm 3$ min, was ever interpreted in terms of the oscillation
66: of a filament. In this paper, we propose a new model to explain
67: this phenomenon by means of 2.5-dimensional MHD numerical
68: simulations. It is found that magnetic reconnection is induced as
69: the coronal loops rooted on both stars inflate and approach each
70: other, which is driven by the differential stellar rotation. The
71: magnetic reconnection is modulated by fast-mode magnetoacoustic
72: waves which are trapped between the surfaces of the two stars, so
73: that the reconnection rate presents a periodic behaviour. With the
74: typical parameters for the binary system, the observed period can
75: be reproduced. We also derive an empirical formula to relate the
76: period of the flaring rate to the coronal temperature and density,
77: as well as the magnetic field.
78: \end{abstract}
79:
80: \begin{keywords}
81: binaries: close -- stars: flare -- stars: oscillations -- waves --
82: methods: numerical
83: \end{keywords}
84:
85: \section{Introduction}
86:
87: The binary system YY Gem has been studied for many years since discovered
88: \citep{Adams20}. It is a double-lined spectroscopic eclipsing binary
89: system which contains two dM1e late-type stars with masses and radii
90: almost identical, and the orbital period being $0.8142822$ d
91: (\citealt{Joy26}; \citealt{vanGent31}). It is of great importance to
92: study the basic relations of such rare
93: late-type binaries (\citealt{Kron52}; \citealt*{Butler96}). Based
94: on some considerations in \citet{Dworak75}, \citet{Brancewicz80}
95: gave the parameters of YY Gem: $M_1 = M_2 = 0.57$ M$_{\odot}$,
96: $R_1 = R_2 = 0.6$ R$_{\odot}$, the separation of the two components $a
97: = 3.83$ R$_{\odot}$.
98:
99:
100: \citet{Chabrier95} demonstrated that the depth of each stellar
101: radiative core is $\sim$ 70 per cent of its radius, and the
102: thickness of the convective zone is $\sim$ 30 per cent. Strong
103: convective motions under the surfaces cause large-scale star spots
104: and huge magnetic field structures. Flare activity on YY Gem was
105: reported for the first time by \citet{Moffett71}. Subsequent
106: investigations have shown that it is one of the most active
107: flaring binaries (\citealt{Moffett74}; \citealt{Doyle85};
108: \citealt{Doyle90a}). Their flaring activities have been studied in
109: multi-wavelengths. Using the Very Large Array, \citet*{Jackson89}
110: presented radio observations; \citet{Stelzer02} studied the
111: simultaneous X-ray spectroscopy of YY Gem with \emph{Chandra} and
112: \emph{XMM-Newton} satellites; While, far UV observations of YY Gem
113: were reported by \citet{Saar03}. The flarings on YY Gem exhibit UV
114: and X-ray emissions which are stronger than those on the Sun
115: \citep{Haisch90}. For example, \citet{Tsikoudi00} observed two
116: large flares, and the integrated X-ray luminosity was estimated to
117: be about 6-8 $\times10^{33}$ erg. They also calculated the ratio
118: of X-ray and bolometric luminosities $L_{X} / L_{Bol}$, which
119: indicated strong magnetic activities and `hot' coronal components.
120:
121: For close binaries with similar masses, the density scale-height
122: is quite large, i.e., the density does not decrease with height
123: from the stellar surfaces as rapidly as in the solar corona.
124: Therefore, the hot plasma between the two stars could emit strong
125: X-ray light. For example, observations revealed large-scale strong
126: X-ray sources between the two stars of the RS CVn binary AR Lac
127: (\citealt{Siarkowski92}; \citealt{Siarkowski96}), as well as of
128: the RS CVn binary TY Pyx (\citealt{Culhane90}; \citealt*{Pres95}).
129: At the same time, it is probable that the two stars are
130: magnetically connected, as illustrated by \citet{Uchida83}.
131: \citet{Uchida85} proposed that the interstellar activities are
132: intimately affected by the differential rotations of both stars.
133: So, there may exist some interesting phenomena in such a system.
134: One of them is the periodic flaring rate, which was reported by
135: \citet{Doyle90a}. Their observation in the \emph{U}-band showed
136: four flares separated in succession with a periodicity of $48 \pm
137: 3$ min during a total observing time of 408 min. The duration of
138: each flare varies from $\sim 20$ min to $\sim 40$ min, and the
139: time-averaged flare luminosity in the \emph{U}-band is 1.25
140: $\times$ $10^{28}$ erg s$^{-1}$. They interpreted this periodicity
141: in terms of filament oscillations. In this paper, we attempt to
142: propose an alternative explanation for the periodicity.
143:
144: Flares on the Sun and dKe-dMe stars are both thought to result
145: from magnetic reconnection. On the Sun, periodic behaviours have
146: also been reported in various reconnection-associated phenomena,
147: e.g., the quasi-periodic modulation of flaring emission
148: (\citealt{Nakariakov06}) and the repetitive appearance of
149: transition region explosive events (\citealt{Chenpf06}). In the
150: former case, magnetic reconnection is modulated by magnetoacoustic
151: waves from a nearby oscillating coronal loop, whereas in the
152: latter, magnetic reconnection is modulated by {\it p}-mode waves.
153:
154: To our knowledge, no efforts in MHD numerical simulations have
155: been devoted to the investigation of flaring phenomenon in binary
156: systems, although hydrodynamic simulations have been done
157: regarding flares in M dwarf stars (e.g. \citealt{Chengcc91}).
158: Using potential field extrapolation, \citet{Uchida85} demonstrated
159: that the evolving surface magnetic field can result in magnetic
160: reconnection in the interstellar space. Here, we perform MHD
161: numerical simulations in order to investigate the magnetic
162: reconnection in a close binary system, with the purpose to clarify
163: which process is likely to be responsible for the periodicity of
164: the flaring rate as reported by \citet{Doyle90a}. The paper is
165: organised as follows. Section 2 describes the basic model, MHD
166: equations, initial and boundary conditions. Section 3 gives the
167: numerically calculated results. Discussions are presented in
168: section 4, which is followed by conclusions in section 5.
169:
170: \section[]{numerical method}
171: \subsection{Problem setup}
172:
173: \begin{figure*}
174: \centering
175: \includegraphics[width=14cm]{fig1.ps}
176: \caption{Sketch of the binary system. For the case of YY Gem,
177: the two stars are identical. Here, $\delta
178: \emph{x}$ is the deviation of the $x$-coordinate from the centre of
179: mass, $r_{1}$ is the distance to the centre of $M_{1}$, and $r_{2}$ is
180: the distance to the centre of $M_{2}$. The simulation box is indicated
181: by the dashed box.
182: }\label{fig1}
183: \end{figure*}
184:
185: The basic model used in our work is shown in Fig. 1. For ordinary
186: binaries, the member stars are usually of different types. As
187: usual, we call the one with larger mass the primary, marked as
188: $M_1$, the other the secondary, marked as $M_2$ ($M_1 > M_2$).
189: Owing to the difference in mass, the point of zero gravity and the
190: centre of mass are located at different positions. The former is
191: closer to the secondary ($M_2$) and the latter is closer to the
192: primary ($M_1$). In the case of YY Gem, we take the parameters
193: given by \citet{Brancewicz80}, i.e., $M_1=M_2$, so that the two
194: points are actually cospatial. As shown in Fig. \ref{fig1}, $a$ is
195: the separation of two stellar centres, and $D=2.63$ R$_{\odot}$ is
196: the distance between the two surfaces along the joint line of the
197: stars.
198:
199: Because of the short orbital period, \citet{Struve59} and
200: \citet{Qian02} suggested that YY Gem should be in synchronized
201: rotation. Therefore, we assume that the angular velocities of both
202: stars have the same value as that of the co-rotating reference
203: system. This means that the two stars rotate almost face to face
204: except the differential rotation. As a result of the differential
205: rotation, persistent shear motion is imposed to the coronal loops
206: whose footpoints are rooted at different latitudes.
207:
208: The MHD processes driven by the differential rotation are
209: three-dimensional in nature. The addition of the two spherical
210: boundaries makes the problem more unfeasible to solve numerically.
211: To simplify the problem without losing the essence of physics, the
212: spherical boundaries are treated as two parallel planes, which are
213: tangent to the spheres at the equators as illustrated with the
214: dotted lines in Fig. \ref{fig1}. The area of interest is $0
215: \leqslant x \leqslant D$ and $-2.5$ R$_{\odot}\leqslant y
216: \leqslant 2.5$ R$_{\odot}$, where the $x$-axis is along the joint
217: line of the binary, the $y$-axis is parallel to the axis of the
218: rotation, and the $z$-axis is perpendicular to the $x-y$ plane,
219: with the origin of the coordinate being located on the equator of
220: $M_1$ star. Note that the vertical size is set to be larger than
221: the stellar radius in order to minimize the influence of the top
222: boundary on the numerical results. Moreover, all quantities are
223: approximated to be invariant along the $z$-axis, so that the
224: problem becomes 2.5-dimensional. In reality, the interface between
225: the magnetic systems of the two stars has a limited extension in
226: the $z$-direction. Therefore, it should be kept in mind that any
227: feature in our numerical results should have a limited size in the
228: $z$-direction.
229:
230: \subsection{MHD equations}
231:
232: We perform 2.5-dimensional $(\partial/\partial z = 0)$ numerical
233: simulations in the Cartesian coordinates as described above.
234: The MHD equations are slightly modified from \citet{Chenpf99a,
235: Chenpf99b} in order to include the gravity from both stars and the
236: centrifugal force in the co-rotating system. For simplicity, heat
237: conduction is not included. The resulting dimensionless MHD
238: equations, which are shown below, are numerically solved by a
239: multistep implicit scheme \citep*{Hu89,chen00}.
240:
241: \begin{equation}%\label{(1)}
242: \frac{\partial \rho}{\partial t} + v_{x} \frac{\partial
243: \rho}{\partial x} + v_{y} \frac{\partial \rho}{\partial
244: y} + \rho \frac{\partial v_{x}}{\partial x} + \rho \frac{\partial
245: v_{y}}{\partial y} = 0,
246: \end{equation}
247:
248: \begin{equation}\label{m2}
249: \frac{\partial v_{x}}{\partial t} + v_{x} \frac{\partial
250: v_{x}}{\partial x} + v_{y} \frac{\partial v_{x}}{\partial
251: y} + \frac{T}{\rho} \frac{\partial \rho}{\partial x} +
252: \frac{\partial T}{\partial x} + \frac{2B_{z}}{\rho \beta_{0}} \frac{\partial B_{z}}
253: {\partial x} + \frac{2}{\rho \beta_{0}} \frac{\partial \psi }
254: {\partial x} \Delta \psi - g - F_{c} = 0,
255: \end{equation}
256:
257: \begin{equation}\label{m3}
258: \frac{\partial v_{y}}{\partial t} +v_{x} \frac{\partial v_{y}}
259: {\partial x} + v_{y}\frac{\partial v_{y}}{\partial y} +
260: \frac{\partial T}{\partial y} + \frac{T}{\rho} \frac{\partial
261: \rho}{\partial y} + \frac{2}{\rho \beta_{0}} \frac{\partial
262: \psi}{\partial y} \Delta \psi + \frac{2B_{z}}{\rho
263: \beta_{0}} \frac{\partial B_{z}}{\partial y}= 0,
264: \end{equation}
265:
266: \begin{equation}%\label{4}
267: \frac{\partial v_{z}}{\partial t} + v_{x} \frac{\partial
268: v_{z}}{\partial x} +v_{y} \frac{\partial v_{z}}{\partial y} +
269: \frac{2}{\rho \beta_{0}} \frac{\partial \psi}{\partial x}
270: \frac{\partial B_{z}}{\partial y} - \frac{2}{\rho \beta_{0}}
271: \frac{\partial \psi}{\partial y} \frac{\partial B_{z}}{\partial
272: x} = 0,
273: \end{equation}
274:
275: \begin{equation}%\label{5}
276: \frac{\partial \psi}{\partial t} + v_{x} \frac{\partial
277: \psi}{\partial x} + v_{y} \frac{\partial \psi}{\partial y} -
278: \frac{1}{R_{m}} \Delta \psi = 0,
279: \end{equation}
280:
281: \begin{equation}%\label{6}
282: \frac{\partial B_{z}}{\partial t} - \frac{\partial v_{z}}{\partial
283: x} \frac{\partial \psi}{\partial y} + \frac{\partial v_{z}}
284: {\partial y} \frac{\partial \psi}{\partial x} + B_{z}
285: \frac{\partial v_{x}}{\partial x} + B_{z} \frac{\partial
286: v_{y}}{\partial y} + v_{x} \frac{\partial B_{z}}{\partial x} +
287: v_{y} \frac{\partial B_{z}}{\partial y} - \frac{\partial}{\partial
288: x} \left(\frac{1}{R_{m}} \frac{\partial B_{z}}{\partial x}\right) -
289: \frac{\partial}{\partial y} \left(\frac{1}{R_{m}} \frac{\partial B_{z}}{\partial
290: y}\right) = 0,
291: \end{equation}
292:
293: \begin{equation}%\label{7}
294: \frac{\partial T}{\partial t} + v_{x} \frac{\partial T}{\partial
295: x} + v_{y} \frac{\partial T}{\partial y} + (\gamma - 1)T
296: \frac{\partial v_{x}}{\partial x} + (\gamma -1)T \frac{\partial
297: v_{y}}{\partial y} - \frac{2(\gamma - 1)}{\rho \beta_{0} R_{m}}
298: \left[\left(\frac{\partial B_{z}}{\partial
299: x}\right)^{2} + \left(\frac{\partial B_{z}}{\partial y}\right)^{2} + (\Delta
300: \psi)^2\right] = 0,
301: \end{equation}
302: \noindent where $\rho$, $v_{x}$, $v_{y}$, $v_{z}$, $\psi$, $B_{z}$
303: and $T$ correspond to the dimensionless density, three components
304: of velocity, magnetic flux function, $z$-component of the magnetic
305: strength, and temperature, respectively, and $\rho_{0}$, $v_{0}$,
306: $\psi_{0}$, $B_{0}$ and $T_{0}$ are the characteristic values of
307: the corresponding parameters, which are used to nondimensionalize
308: the equations. The characteristic velocity $v_{0}$ is defined as
309: the isothermal sound speed $v_{0} = \sqrt{RT_0}$, where $R$ is the
310: gas constant for the fully ionised hydrogen and the length scale
311: $L_{0}$ is set to be 0.5 R$_{\odot}$. The dimensionless size of
312: the simulation box is then $0\leq |y| \leq 5$ and $0\leq x \leq
313: 5.26$. Since the flux function and the magnetic field are related
314: by
315: \begin{equation}%\label{8}
316: \textbf{\emph{B}} = \nabla \times (\psi \textbf{\emph{e}}_{z})
317: + B_{z} \textbf{\emph{e}}_{z},
318: \end{equation}
319: where $\textbf{\emph{e}}_{z}$ is the unit vector in the
320: $\emph{z}$-axis, we have $\psi_{0} = B_{0} L_{0}$.
321:
322: In equation (\ref{m2}), $g$ is the total acceleration of gravity
323: contributed by both stars,
324: \begin{equation}%\label{9}
325: g = \frac{G }{v_{0}^{2} L_{0}} (\frac{M_{2}}{r_{2}^{2}} - \frac{M_{1}}{r_{1}^{2}}),
326: \end{equation}
327: where $M_{1}$ and $M_{2}$ are the masses of the two stars, $r_1$
328: and $r_2$ are indicated in Fig. \ref{fig1} (note that such an
329: expression is an extension of the gravity along the $x$-axis,
330: which is valid since we are interested in the processes near the
331: $x$-axis). In the same equation, $F_{c}$ represents the
332: centrifugal force,
333: \begin{equation}%\label{10}
334: F_{c} = \frac{G (M_{1} + M_{2}) L_{0}^{2}}{v_{0}^{2}
335: a^3} \delta x,
336: \end{equation}
337: where $a$ is the distance between the two stars, $\delta \emph{x}$
338: is the deviation of the $x$-coordinate from the centre of mass.
339: The plasma beta, $\beta_{0}$, and the magnetic Reynolds number,
340: $R_m$, are expressed as
341: \begin{equation}%\label{11}
342: \beta_{0} = \frac{2\mu_{0}\rho_{0}v_{0}^{2}}{B_{0}^{2}},
343: \end{equation}
344: \begin{equation}%\label{12}
345: R_{m} = \frac{\mu_{0}v_{0}L_{0}}{\eta},
346: \end{equation}
347: where $\mu_{0}$ is the magnetic permeability and $\eta$ is the
348: anomalous resistivity. Similarly to \citet{Chenpf06}, $\eta$ is
349: chosen to be a function of the current density, $j$,
350: \begin{equation}%\label{16}
351: \eta = \left\{
352: \begin{array}{cc}
353: \eta_{0} \min(1, j/j_{c} - 1), & j \geqslant j_{c}, \\
354: 0, & j < j_{c}, \\
355: \end{array}\right.
356: \end{equation}
357: where $j_{c}$ is the critical current density.
358:
359: The time unit used in the description of the numerical results is the
360: Alfv\'{e}n time-scale, $\tau_{A} = L_{0}/v_{A}$, where $v_{A}$ is
361: Alfv\'{e}n speed, i.e.,
362: \begin{equation}%\label{13}
363: v_A = \frac{\psi_{0}}{L_{0}\sqrt{\mu_{0}\rho_{0}}}.
364: \end{equation}
365:
366: As \citet{Tsikoudi00} pointed out, the corona of YY Gem is very
367: hot, $T_0$ can be taken to be several million kelvin. The number
368: density ($n=\rho/m_p$) of the stellar corona is in the range of
369: $10^{14}$ -- $10^{18}$ m$^{-3}$ (\citealt{Monsignori94};
370: \citealt{Mewe95}; \citealt{Schmitt96}; \citealt{Schrijver00};
371: \citealt{Ness02}). The typical strength of the general magnetic
372: field near the stellar surface is tens of Gauss. Therefore, in our
373: standard case, which is called case A2 later, the following
374: characteristic values are adopted: $n_0=10^{15}$ m$^{-3}$,
375: $T_0=1.9\times 10^6$ K, $\beta_0$=0.01 (hence $B_0$=36 G), and
376: $j_c=10$. The corresponding time-scale is $\tau_A=139$ s.
377:
378: \subsection{Initial and boundary conditions}
379:
380: \begin{figure}
381: \centering
382: \includegraphics[width=9cm,height=5cm]{fig2.ps}\\
383: \caption{Initial configuration of the magnetic field (solid lines) and
384: the distribution of the dimensionless density (gray scale).}\label{fig2}
385: \end{figure}
386:
387: Initially, the isothermal plasma is at rest, i.e.,
388: $v_x=v_y=v_z=0$, and $T=1$. The density distribution is obtained
389: by numerically calculating the hydrostatic equation
390: \begin{equation}%\label{17}
391: \frac{1}{\rho} \frac{\textrm{d}\rho}{\textrm{d}x} = g(x) + F_{c}(x).
392: \end{equation}
393:
394: The initial magnetic field in this work is assumed to be
395: potential, which is produced by two parallel line currents lying
396: below the two surfaces, i.e.,
397:
398: \begin{equation}\label{mag}
399: \psi =-\ln(\sqrt{(x - x_{1})^2 + y^2}) - \ln(\sqrt{(x -
400: x_{2})^2 + y^2}),
401: \end{equation}
402: where $x_1$ and $x_2$ are the $x$-coordinates of these two
403: currents. We take $x_1 = -1$ and $x_2 = D/L_0 + 1$ in our work.
404: The resulting density distribution and the magnetic configuration
405: are plotted in Fig. \ref{fig2}, where the magnetic field is
406: similar to the configuration adopted in \citet{Ferreira98}. The
407: magnetic configuration is taken in such a way that a magnetic null
408: point exists between the antiparallel magnetic loops. So, whenever
409: the closed loops bulge, a current sheet forms natually, and
410: magnetic reconnection is ready to commence (if the polarity of one
411: magnetic system expressed in equation (\ref{mag}) is flipped, the
412: two loop systems become to repel each other, and no reconnection
413: can happen). Similar magnetic connectivity with a null point was
414: also computed for the active RS CVn system (\citealt{Uchida83};
415: \citealt{Uchida85}; \citealt{Beasley00}).
416:
417: Owing to the symmetry, calculations are done only in the upper
418: half of the simulation area, i.e., $0 \leqslant x \leqslant D$ and
419: $0\leqslant y \leqslant 2.5$ R$_{\odot}$. The calculation domain
420: is discretised into 501 uniform grid points along the $x$-axis and
421: 141 nonuniform grid points along the $y$-axis, with more grid
422: points concentrated near the $x$-axis. The top side of the
423: simulation box ($y=5$) is treated as an open boundary, and the
424: bottom one ($y=0$) is a symmetry boundary. The right and left,
425: which correspond to the surface of the binary stars, are line-tied
426: boundaries, where all quantities are fixed except that a typical
427: differential rotation is imposed at each surface in the following
428: form:
429:
430: \begin{equation}%\label{19}
431: v_z|_{\textrm{left}} = \left\{
432: \begin{array}{cc}
433: - \Omega R(1-\alpha \sin^{2}i) \sqrt{1-\sin^{2}i} + \Omega R ,& i \leqslant \pi/2,\\
434: \Omega R,& i > \pi/2, \\
435: \end{array}\right.
436: \end{equation}
437:
438: \begin{equation}%\label{20}
439: v_z|_{\textrm{right}} = \left\{
440: \begin{array}{cc}
441: \Omega R(1-\alpha \sin^{2}i)\sqrt{1-\sin^{2}i} - \Omega R ,& i \leqslant \pi/2,\\
442: - \Omega R,& i > \pi/2, \\
443: \end{array}\right.
444: \end{equation}
445: where $\Omega$ is the angular velocity of the binary, $\alpha$ is
446: the differential rotation rate, and $i$, the latitude, is related
447: to the coordinate $y$ by $y=R\sin i$. Beyond $y=R$, $v_z$ is set
448: to be the value at $y=R$. In this paper, $\alpha$ is set to be
449: 0.5. As we will discuss later, the value of $\alpha$ does not
450: affect the main results in our numerical simulations.
451:
452: \section{numerical results}
453:
454: \subsection{General evolution}
455:
456: \begin{figure*}
457: \centering
458: \includegraphics[width=6cm,height=15cm]{fig3.ps}\\
459: \caption{Magnetic field lines (solid lines), velocities (arrows),
460: and temperatures (colour) at five selected times in case A2.}
461: \label{fig3}
462: \end{figure*}
463:
464: The distributions of the magnetic field, the temperature, and the
465: velocity field at several selected times in case A2 are shown in
466: Fig. \ref{fig3}. As the shear motion is imposed at the left and
467: right boundaries, footpoints of the coronal loops are dragged to
468: move in the $z$-direction, which increases the local magnetic
469: pressure. As a result, both the left and the right magnetic loop
470: systems begin to inflate and approach each other due to the
471: increasing pressures near the roots of the loop systems (see the
472: snapshot at $t=83$ $\tau_A$). Such an inflation of magnetic loops
473: due to shear motions was demonstrated on the Sun \citep{Barnes72}.
474: The two loop systems collide near the magnetic null point, by
475: which a current sheet forms. When the increasing current density
476: exceeds the prescribed threshold ($j_c$), anomalous resistivity is
477: excited, which triggers the occurrence of magnetic reconnection
478: between the two loop systems (see the snapshot at $t=109$
479: $\tau_A$). The temporal evolution of the reconnection rate, $R_r$,
480: which is defined as $\textrm{d}\psi/\textrm{d}t$ at the magnetic
481: null point, is shown in Fig. \ref{A2rate}. It is seen that the
482: reconnection rate presents a periodic behaviour. With the wavelet
483: analysis, the period is found to be $54 \pm 6$ min. Note that,
484: with the 2.5D approximation in our model, the reconnection between
485: the two closed magnetic systems, which are rooted on the surfaces
486: of the binary stars, will go on as long as the differential
487: rotation continues. In reality, the differential rotation of the
488: two spherical stellar surfaces would take the two closed magnetic
489: systems apart, which would cease the reconnection after a certain
490: time. The 2.5D model is valid here since we are interested in the
491: magnetic reconnection only during first several periods.
492:
493: \begin{figure}
494: \centering
495: \includegraphics[width=8cm]{fig4.ps}\\
496: \caption{Temporal evolution of the reconnection rate in case A2, where
497: $R_r$ is the reconnection rate, and $\tau_A=139$ s.}\label{A2rate}
498: \end{figure}
499:
500: In order to compare the numerical result with the \emph{U}-band
501: observations made by \citet{Doyle90a}, we synthesised the
502: \emph{UV} emission in the wave-band of $3650\pm10${\AA} from our
503: numerical results with the help of the CHIANTI database
504: (\citealt{dere97}; \citealt{land06}; see \citealt{Chenpf06} for
505: details). The calculated \emph{UV} light-curve of the whole
506: simulation area is plotted in the left panel of Fig. 5, and its
507: wavelet spectrum is presented in the right panel. Note that since
508: the reconnection rate and the \emph{UV} emission both are small in
509: the first 60 $\tau_A$, we displayed the result after $t=60
510: \tau_A$. It is seen that the \emph{UV} flux does show a
511: periodicity, and the period is centred at 52 min, very close to
512: that of the reconnection rate, i.e., 54 min.
513:
514: \begin{figure*}
515: \centering
516: \includegraphics[width=14cm]{fig5.ps}\\
517: \caption{Left panel: Temporal evolution of the flux in the \emph{U}-band
518: , which is derived by using the CHIANTI database. Right
519: panel: The wavelet spectrum corresponding to the \emph{U}-band
520: flux, which has a period centred around about 52 min.
521: }\label{urad}
522: \end{figure*}
523:
524: \subsection{Parameter survey}
525:
526: In order to investigate how the choice of the characteristic
527: values of various parameters may affect the results, we have made
528: numerical simulations with different characteristic values, which
529: are listed in Table \ref{tab1}. In cases A1--A5, only $T_0$ is
530: changed. Similarly, cases B1--B5 are for $\beta_0$ dependence,
531: cases C1--C3 for $\rho_0$, and cases D1--D3 for $j_c$,
532: respectively. In all cases, $L_0$ is fixed to be 0.5 R$_{\odot}$.
533: Numerical simulations indicate that, similar to case A2 as
534: described in the previous subsection, all other cases present
535: periodic behaviours. The corresponding period of each case is
536: displayed in the last column of Table \ref{tab1}. It is found
537: that, with $T_0$, $\beta_0$, $\rho_0$, and $j_c$ as independent
538: parameters, the period of the flaring is strongly dependent on
539: $T_0$ and $\beta_0$, and an increasing $T_0$ or decreasing
540: $\beta_0$ results in a decreasing period. When $\rho_0$ increases,
541: the magnetic field $B_0$ increases as well in order to keep
542: $\beta_0$ constant (correspondingly, the Alfv\'en speed is constant).
543: As a result, the period of the reconnection rate does not
544: change.It is also seen from Table \ref{tab1} that $j_c$ has little
545: effect on the period, which implies that the magnetic reconnection
546: is modulated by some external process.
547:
548:
549: \begin{table}
550: \centering
551: \caption{Parameters used in different simulation cases.}
552: \begin{tabular}{|c|c|c|c|c|c|}
553: \hline
554: % after \\: \hline or \cline{col1-col2} \cline{col3-col4} ...
555: Models & $T_{0}$ ($\times 10^{6}$ K) & $\beta_{0}$ & $\lg n_{0}$/m$^{-3}$ & $j_{c}$ & $P$ (min) \\
556: % \footnote{unit is m^{-3}}\\
557: \hline
558: A1 & 1.5 & 0.01 & 15 & 10.0 & 70 $\pm$ 9 \\
559: A2 & 1.9 & 0.01 & 15 & 10.0 & 54 $\pm$ 6 \\
560: A3 & 2.5 & 0.01 & 15 & 10.0 & 41 $\pm$ 4 \\
561: A4 & 3.0 & 0.01 & 15 & 10.0 & 35 $\pm$ 3 \\
562: A5 & 4.0 & 0.01 & 15 & 10.0 & 31 $\pm$ 3 \\
563: \hline
564: B1 & 2.0 & 0.05 & 15 & 10.0 & 96 $\pm$ 15\\
565: B2 & 2.0 & 0.025 & 15 & 10.0 & 78 $\pm$ 10\\
566: B3 & 2.0 & 0.01 & 15 & 10.0 & 52 $\pm$ 6 \\
567: B4 & 2.0 & 0.0075 & 15 & 10.0 & 47 $\pm$ 4 \\
568: B5 & 2.0 & 0.005 & 15 & 10.0 & 38 $\pm$ 4 \\
569: \hline
570: C1 & 2.0 & 0.01 & 14 & 10.0 & 52 $\pm$ 6 \\
571: C2 & 2.0 & 0.01 & 15 & 10.0 & 52 $\pm$ 6 \\
572: C3 & 2.0 & 0.01 & 16 & 10.0 & 52 $\pm$ 6 \\
573: \hline
574: D1 & 2.0 & 0.01 & 15 & 1.0 & 52 $\pm$ 6 \\
575: D2 & 2.0 & 0.01 & 15 & 5.0 & 52 $\pm$ 6 \\
576: D3 & 2.0 & 0.01 & 15 & 10.0 & 52 $\pm$ 6 \\
577: \hline
578: \end{tabular}\label{tab1}
579: \end{table}
580:
581: Besides the four parameters mentioned above, we also examine other
582: parameters related to the binary itself. In case A2, by changing
583: the values of $\alpha$ ($\alpha = 0.2$, $0.3$, $0.4$, $0.5$,
584: $0.6$, $0.7$, $0.8$), we find that the results keep invariant,
585: i.e., $P=54 \pm 6$ min. In other words, the flaring period has
586: nothing to do with the speed of the differential rotation. As
587: mentioned above, the mass and the size of YY Gem are taken from
588: \citet{Brancewicz80}. However, other authors derived slightly
589: different values (\citealt{Leung78}; \citealt{Chabrier95};
590: \citealt{Torres02}), e.g., in \citet{Leung78}, $M_{1} =0.62$
591: M$_{\odot}$, $M_{2} = 0.57$ M$_{\odot}$, $R_{1} = 0.66$
592: R$_{\odot}$, $R_{2} = 0.58$ R$_{\odot}$, the separation of two
593: components $a = 3.9$ R$_{\odot}$, and the orbital period $P =
594: 0.82$ d. Through test calculations, we find little change in the
595: results.
596:
597: \section{Discussions}
598:
599: \subsection{What modulates the magnetic reconnection?}
600: In order to understand how the magnetic reconnection is modulated,
601: the temporal evolution of $v_{x}$ along the $x$-axis is shown in
602: Fig. \ref{vxgray}, where the value of $v_x$ is indicated by the
603: grey scales. As the reconnection rate $R_r$ is equivalent to
604: $v_xB_y$ along the $x$-axis, it is expected to see that $v_x$ at a
605: given point also shows an oscillatory evolution, with the same
606: period as $R_r$, i.e., 54 min. Furthermore, a wave pattern is
607: discerned in Fig. \ref{vxgray}, where the waves are seen to travel
608: back and forth in the $x$-direction, bounded by the surfaces of
609: the two stars. Since the waves propagate perpendicular to the
610: magnetic field lines, we postulate that they are fast-mode
611: magnetoacoustic waves, and that the periodic magnetic reconnection
612: in the numerical results is modulated by such waves.
613:
614: \begin{figure}
615: \centering
616: \includegraphics[width=8cm,height=6cm]{fig6.ps}\\
617: \caption{Temporal evolution of $v_{x}$ along the $x$-axis.}\label{vxgray}
618: \end{figure}
619:
620: \begin{figure}
621: \centering
622: \includegraphics[width=8cm]{fig7.ps}\\
623: \caption{Temporal evolution of reconnection rate in the case that
624: the horizontal size of the simulation box is cut down.
625: }\label{d_cut}
626: \end{figure}
627:
628: To verify that the magnetic reconnection is modulated by
629: magnetoacoustic waves bouncing back and forth between the two
630: stellar surfaces, we shift the right boundary of case A2 leftward
631: and the left boundary rightward by 0.8 simultaneously, by which
632: the separation between the two stellar surfaces is cut down from
633: 5.26 to 3.66, i.e., a decrease with a ratio of 30 per cent. The
634: corresponding evolution of the reconnection rate is plotted in
635: Fig. \ref{d_cut}. Note that the positive $R_r$ means that the
636: field lines in the two loop systems are reconnecting, whereas the
637: negative $R_r$ means that the transverse field lines are
638: reconnecting.
639:
640: It is found that the resulting period of magnetic reconnection
641: becomes 34 $\pm$ 6 min. This means that the period decreases by 37
642: per cent, a ratio close to that of the decrease of the separation
643: between the two stellar surfaces. This implies that the magnetic
644: reconnection is modulated by waves trapped between the two stellar
645: surfaces.
646:
647: To verify that the travelling waves are fast-mode waves, here we
648: calculate the trajectory of fast-mode waves propagating from the
649: reconnection point ($x=2.63$), through the right boundary
650: ($x=5.26$) and then back to the reconnection point along the
651: $x$-axis. For a given number $n$, the travelling time ($T$) and
652: the position of the wave front ($x$) are determined by the
653: following equations
654:
655: \begin{equation}\label{tw1}
656: T =\sum_{i\geqslant 1}^{n}\Delta x/v_f, \\
657: x-2.63=\sum_{i\geqslant 1}^{n}\Delta x,
658: \end{equation}
659: \noindent where $\Delta x$ is the uniform grid spacing in the
660: $x$-direction, and $v_f$ is the fast-mode wave velocity. The
661: procedure is described as follows: Starting from the reconnection
662: point ($x$=2.63), the wave front propagates rightward by $\Delta
663: x$ each step, the time interval $\Delta t$ is equal to $\Delta x$
664: divided by $v_f$, so the total travelling time is the sum of
665: $\Delta t$, while the position is the sum of $\Delta x$. Note that
666: since both the magnetic field and the plasma temperature are
667: changing with time, we update the magnetic and thermal parameters
668: every 1 $\tau_A$ when calculating $v_f$. Besides, when the wave
669: reaches the right boundary and then bounces back, $\Delta x$ takes
670: a negative value. After the wave front returns to the reconnection
671: point, the wave should propagate to the left half area of the
672: simulation box. Considering the symmetry of the problem, i.e.,
673: there is a symmetric wave propagating rightward from the left half
674: area, we repeat the above calculation in the right half region.
675:
676: \begin{figure}
677: \centering
678: \includegraphics[width=8cm]{fig8.ps}\\
679: \caption{Temporal evolution of the wave front. Only the right half is
680: plotted owing to the symmetry of the YY Gem system.
681: }\label{front}
682: \end{figure}
683:
684: Figure. \ref{front} shows the temporal evolution of the position
685: of the wave front in the right half region only. It is found that
686: the trajectory of the fast-mode wave, from the reconnection point
687: to the right (or left) boundary and then back to the reconnection
688: point, is quasi-periodic, with a period of $\sim23\pm2$ $\tau_A$,
689: i.e., $\sim 53 \pm 5$ min, very close to the period of the
690: magnetic reconnection rate presented in Section 3. Therefore, we
691: are convinced that the magnetic reconnection is modulated by
692: fast-mode waves. The underlying process can be understood as
693: follows: When symmetric waves, introduced either by the shear
694: motions of the stellar surfaces or by the magnetic reconnection,
695: collide near the reconnection point, the reconnection rate is
696: enhanced to form a peak. After the collision, the two waves
697: propagate to the right and left boundaries and are bounced back,
698: respectively. When they collide again near the reconnection point,
699: the reconnection rate is enhanced again. Therefore, the period of
700: the magnetic reconnection is determined by the travelling time of
701: fast-mode waves from the reconnection point to the right (or left)
702: boundary and then back to the reconnection point.
703:
704: \subsection{Empirical formula for the period}
705:
706: As shown in Section 3, among the selected four parameters, the
707: period of the reconnection rate ($P$) is explicitly dependent on
708: $T_0$ and $\beta_0$. The underlying process can be understood as
709: follows: When $T_0$ increases while keeping $\beta_0$ and $\rho_0$
710: constant, both the sound speed and the Alfv\'en speed increase,
711: which leads to a larger fast-mode wave speed. As a result, the
712: period of the reconnection rate decreases; When $\beta_0$
713: decreases while keeping $T_0$ and $\rho_0$ constant, the sound
714: speed does not change, while the Alfv\'en speed increases, which
715: also leads to a larger fast-mode wave speed and a shorter period;
716: When $\rho_0$ increases while keeping $\beta_0$ and $T_0$
717: constant, both the sound and the Alfv\'en speed do not change (the
718: Alfv\'en speed does not change since the magnetic field $B_0$
719: increases). As a result, the period keeps constant.
720:
721: In order to derive an empirical formula to relate $P$ to $T_0$ and
722: $\beta_0$, we simulate a series of cases with different $T_0$ and
723: $\beta_0$. The corresponding variations of $P$ with $T_0$ and
724: $\beta_0$ are displayed by the squares in Fig. \ref{P-T}. It is
725: seen that $\lg P$ is almost a linear function of $\lg T_0$ and
726: $\lg \beta_0$, which means that $P$ scales with $T_0$ and
727: $\beta_0$ as $P \propto T_0^{\gamma_1} \beta_0^{\gamma_2}$. By
728: fitting the data points with the solid lines in Fig. \ref{P-T}, we
729: obtain $\gamma_1=-0.91$ and $\gamma_2=0.42$. Therefore, we have $P
730: \propto T_0 ^{-0.91}\beta_0^{0.42}$, i.e.,
731:
732: \begin{equation}\label{emfor}
733: P \sim \rho_0^{0.42} T_0^{-0.49} B_0^{-0.84}.
734: \end{equation}
735:
736: \begin{figure}
737: \centering
738: \includegraphics[width=8cm]{fig9a.ps}\\
739: \includegraphics[width=8cm]{fig9b.ps}\\
740: \caption{Upper panel: Relation between the period ($P$) and the
741: temperature ($T$); Lower panel: Relation between the period
742: ($P$) and $\beta_0$. The solid lines are the corresponding
743: linear fitting of the data points.
744: }\label{P-T}
745: \end{figure}
746:
747: According to our postulation that the magnetic reconnection is
748: modulated by fast-mode waves, the period $P$ should be
749: proportional to $1/v_f$, where $v_f^2=c_s^2+v_A^2$, $c_s^2=\gamma
750: RT_0$, and $v_A^2=B_0^2/(\mu_0 \rho_0)$. In the extreme case of
751: zero magnetic field, we have $P\propto T_0^{-0.5}$; In the case of
752: extremely strong magnetic field, we have $P\propto
753: \rho_0^{0.5}B_0^{-1}$. The slight deviation of the power indices
754: in our numerical results from those in the two extreme cases
755: further implies that it is the fast-mode waves that modulate the
756: magnetic reconnection. Based on this scaling law, the observed
757: period of the flaring rate in YY Gem, i.e., 48 min, corresponds to
758: a characteristic magnetic field of 41 G, which is the value of the
759: magnetic field in low corona, providing the number density of the
760: plasma in the inter-binary space is about $n_0 \sim 10^{15}$
761: m$^{-3}$.
762:
763: \subsection{Other points}
764:
765: The periodic flaring of YY Gem, which was found by
766: \citet{Doyle90a}, was not confirmed in \citet{Stelzer02}. The
767: discrepancy might be attributed to two factors: (1) Probably the
768: two observations were made at different phases of the stellar
769: magnetic cycle. In order to have the large-scale magnetic
770: reconnection in the inter-binary space, the magnetic configuration
771: of each star should be close to the bipolar type, with
772: interconnecting magnetic field lines linking the polar regions of
773: the binary stars, as shown by Fig. \ref{fig2}. Such a magnetic
774: configuration often appears near activity minimum; (2) The
775: periodic flares observed by \citet{Doyle90a} and confirmed by our
776: numerical simulations, are produced in the inter-binary space,
777: therefore, are large-scale in nature. Without doubt, there should
778: be small-scale flares that occur near the stellar surface and are
779: due to the reconnection of the magnetic fields of an individual
780: star. The flares observed by \citet{Stelzer02} probably correspond
781: to these small-scale flares.
782:
783: \section{Conclusions}
784:
785: Observations indicated that there may exist periodicity in the
786: occurrence of strong flares from YY Gem. Using 2.5-dimensional MHD
787: numerical simulations, we confirmed the periodicity of the flaring
788: in the inter-binary space of YY Gem. On the basis of the analysis
789: of the simulation results, the following conclusions can be drawn:
790:
791: 1. Owing to the differential rotation of the stellar surface, coronal
792: loops in each star inflate and approach each other. The resulting
793: magnetic reconnection generates the inter-binary flares, which are
794: large-scale in nature. Fast magnetoacoustic waves, which are trapped
795: in the inter-binary space between the surfaces of the two binary
796: components, modulate the magnetic reconnection, producing the
797: periodic behaviour of the flaring rate.
798:
799: 2. The period scales with the typical coronal density ($\rho_0$),
800: temperature ($T_{0}$), and the coronal magnetic field ($B_0$)
801: as $P\sim \rho_0^{0.42} T_0^{-0.49} B_0^{-0.84}$.
802:
803: \section*{Acknowledgments}
804: We thank the referee for constructive comments that led to an
805: improvement of the paper. We are also grateful to C. Fang, X. Y.
806: Xu, M. Jin, and Q. M. Zhang for their helpful suggestions. This
807: work was supported by the Chinese foundations NCET-04-0445, FANEDD
808: (200226), NSFC under grants 10025315, 10221001, 10333040,
809: 10403003, and 10673004, and by NKBRSF under grant 2006CB806302.
810:
811: %\bibliographystyle{mn2e}
812: %\bibliography{ref}
813:
814: \begin{thebibliography}{}
815:
816: \bibitem[\protect\citeauthoryear{{Adams} \& {Joy}}{{Adams} \&
817: {Joy}}{1920}]{Adams20}
818: {Adams} W.~S., {Joy} A.~H., 1920, PASP., 32, 158
819:
820: \bibitem[\protect\citeauthoryear{{Barnes} \& {Sturrock}}{{Barnes} \&
821: {Sturrock}}{1972}]{Barnes72}{Barnes} C.~W., {Sturrock} P.~A.,
822: 1972, ApJ, 174, 659
823:
824: \bibitem[\protect\citeauthoryear{{Beasley} \& {G{\"u}del}}{{Beasley} \&
825: {G{\"u}del}}{2000}]{Beasley00}
826: {Beasley} A.~J., {G{\"u}del} M., 2000, ApJ, 529, 961
827:
828: \bibitem[\protect\citeauthoryear{{Brancewicz} \& {Dworak}}{{Brancewicz} \&
829: {Dworak}}{1980}]{Brancewicz80}
830: {Brancewicz} H.~K., {Dworak} T.~Z., 1980, Acta Astronomica, 30,
831: 501
832:
833: \bibitem[\protect\citeauthoryear{{Butler}, {Doyle} \& {Budding}}{{Butler}
834: et~al.}{1996}]{Butler96}
835: {Butler} C.~J., {Doyle} J.~G., {Budding} E., 1996, in
836: {Pallavicini} R.,
837: {Dupree} A.~K., eds, Cool Stars, Stellar Systems, and the Sun Vol.~109 of
838: Astronomical Society of the Pacific Conference Series, {Surface
839: inhomogeneities on YY Geminorum}.
840: p 589
841:
842: \bibitem[\protect\citeauthoryear{{Chabrier} \& {Baraffe}}{{Chabrier} \&
843: {Baraffe}}{1995}]{Chabrier95}
844: {Chabrier} G., {Baraffe} I., 1995, ApJ, 451, L29
845:
846: \bibitem[\protect\citeauthoryear{{Chen}, {Fang}, {Ding} \& {Tang}}{{Chen}
847: et~al.}{1999a}]{Chenpf99a}
848: {Chen} P.~F., {Fang} C., {Ding} M.~D., {Tang} Y.~H., 1999a,
849: ApJ, 520, 853
850:
851: \bibitem[\protect\citeauthoryear{{Chen}, {Fang} \& {Hu}}{{Chen} et al.}{2000}]{chen00}
852: Chen, P. F., Fang, C., Hu, Y. Q., 2000, Chin. Sci.
853: Bull., 45, 798
854:
855: \bibitem[\protect\citeauthoryear{{Chen}, {Fang}, {Tang} \& {Ding}}{{Chen}
856: et~al.}{1999b}]{Chenpf99b}
857: {Chen} P.~F., {Fang} C., {Tang} Y.~H., {Ding} M.~D., 1999b,
858: ApJ, 513, 516
859:
860: \bibitem[\protect\citeauthoryear{{Chen} \& {Priest}}{{Chen} \&
861: {Priest}}{2006}]{Chenpf06}
862: {Chen} P.~F., {Priest} E.~R., 2006, Sol. Phys., 238, 313
863:
864: \bibitem[\protect\citeauthoryear{{Cheng} \& {Pallavicini}}{{Cheng} \&
865: {Pallavicini}}{1991}]{Chengcc91}
866: {Cheng} C.-C., {Pallavicini} R., 1991, ApJ, 381, 234
867:
868: \bibitem[\protect\citeauthoryear{{Culhane}, {White}, {Parmar} \&
869: {Shafer}}{{Culhane} et~al.}{1990}]{Culhane90}
870: {Culhane} J.~L., {White} N.~E., {Parmar} A.~N., {Shafer}
871: R.~A., 1990, MNRAS, 243, 424
872:
873: \bibitem[\protect\citeauthoryear{Dere et al.}{1997}]{dere97}
874: {Dere}, K. P., {Landi}, E., {Mason}, H.E. et al.,
875: 1997, Astron. Astrophys. Suppl., 125, 149
876:
877: \bibitem[\protect\citeauthoryear{{Doyle} \& {Butler}}{{Doyle} \&
878: {Butler}}{1985}]{Doyle85}
879: {Doyle} J.~G., {Butler} C.~J., 1985, Nat, 313, 378
880:
881: \bibitem[\protect\citeauthoryear{{Doyle}, {Butler}, {van den Oord} \&
882: {Kiang}}{{Doyle} et~al.}{1990}]{Doyle90a}
883: {Doyle} J.~G., {Butler} C.~J., {van den Oord} G.~H.~J., {Kiang}
884: T., 1990, A\&A, 232, 83
885:
886: \bibitem[\protect\citeauthoryear{{Doyle} \& {Mathioudakis}}{{Doyle} \&
887: {Mathioudakis}}{1990}]{Doyle90b}
888: {Doyle} J.~G., {Mathioudakis} M., 1990, A\&A, 227, 130
889:
890: \bibitem[\protect\citeauthoryear{{Dworak}}{{Dworak}}{1975}]{Dworak75}
891: {Dworak} T.~Z., 1975, Acta Astronomica, 25, 383
892:
893: \bibitem[\protect\citeauthoryear{{Ferreira}}{{Ferreira}}{1998}]{Ferreira98}
894: {Ferreira} J.~M., 1998, A\&A, 335, 248
895:
896: \bibitem[\protect\citeauthoryear{{Haisch}, {Schmitt}, {Rodono} \& {Gibson}}
897: {{Haisch} et~al.}{1990}]{Haisch90}
898: {Haisch} B.~M., {Schmitt} J.~H.~M.~M., {Rodono} M., {Gibson} D.~M., 1990, A\&A,
899: 230, 419
900:
901: \bibitem[\protect\citeauthoryear{{Hu}}{{Hu}}{1989}]{Hu89}
902: {Hu} Y.~Q., 1989, J. Comput. Phys., 84, 441
903:
904: \bibitem[\protect\citeauthoryear{{Jackson}, {Kundu} \& {White}}{{Jackson}
905: et~al.}{1989}]{Jackson89}
906: {Jackson} P.~D., {Kundu} M.~R., {White} S.~M., 1989, A\&A,
907: 210, 284
908:
909: \bibitem[\protect\citeauthoryear{{Joy} \& {Sanford}}{{Joy} \&
910: {Sanford}}{1926}]{Joy26}
911: {Joy} A.~H., {Sanford} R.~F., 1926, ApJ, 64, 250
912:
913: \bibitem[\protect\citeauthoryear{{Kron}}{{Kron}}{1952}]{Kron52}
914: {Kron} G.~E., 1952, ApJ, 115, 301
915:
916: \bibitem[\protect\citeauthoryear{Landi et al.}{2006}]{land06}
917: {Landi} E., {Del Zanna}, G., {Young}, P.R. et al.,
918: 2006, ApJS, 162, 261
919:
920: \bibitem[\protect\citeauthoryear{{Leung} \& {Schneider}}{{Leung} \&
921: {Schneider}}{1978}]{Leung78}
922: {Leung} K.-C., {Schneider} D.~P., 1978, AJ, 83, 618
923:
924: \bibitem[\protect\citeauthoryear{{Mewe}, {Kaastra}, {Schrijver}, {van den Oord}
925: \& {Alkemade}}{{Mewe} et~al.}{1995}]{Mewe95}
926: {Mewe} R., {Kaastra} J.~S., {Schrijver} C.~J., {van den Oord}
927: G.~H.~J., {Alkemade} F.~J.~M., 1995, A\&A, 296, 477
928:
929: \bibitem[\protect\citeauthoryear{{Moffett}}{{Moffett}}{1974}]{Moffett74}
930: {Moffett} T.~J., 1974, ApJS, 29, 1
931:
932: \bibitem[\protect\citeauthoryear{{Moffett} \& {Bopp}}{{Moffett} \&
933: {Bopp}}{1971}]{Moffett71}
934: {Moffett} T.~J., {Bopp} B.~W., 1971, ApJ, 168, L117
935:
936: \bibitem[\protect\citeauthoryear{{Monsignori-Fossi} \&
937: {Landini}}{{Monsignori-Fossi} \& {Landini}}{1994}]{Monsignori94}
938: {Monsignori-Fossi} B.~C., {Landini} M., 1994, A\&A, 284, 900
939:
940: \bibitem[\protect\citeauthoryear{Nakariakov et
941: al.}{2006}]{Nakariakov06} Nakariakov V.~M., Foullon C., Verwichte
942: E., Young N.~P., 2006, A\&A, 452, 343
943:
944: \bibitem[\protect\citeauthoryear{{Ness}, {Schmitt}, {Burwitz}, {Mewe},
945: {Raassen}, {van der Meer}, {Predehl} \& {Brinkman}}{{Ness}
946: et~al.}{2002}]{Ness02}
947: {Ness} J.-U., {Schmitt} J.~H.~M.~M., {Burwitz} V., {Mewe} R.,
948: {Raassen}
949: A.~J.~J., {van der Meer} R.~L.~J., {Predehl} P., {Brinkman} A.~C.,
950: 2002, A\&A, 394, 911
951:
952: \bibitem[\protect\citeauthoryear{{Pres}, {Siarkowski} \&
953: {Sylwester}}{{Pres} et~al.}{1995}]{Pres95}{Pres} P., {Siarkowski}
954: M., {Sylwester} J., 1995, MNRAS, 275, 43
955:
956: \bibitem[\protect\citeauthoryear{{Qian}, {Liu}, {Tan} \&
957: {Soonthornthum}}{{Qian} et~al.}{2002}]{Qian02}
958: {Qian} S.~B., {Liu} D.~L., {Tan} W.~L., {Soonthornthum} B.,
959: 2002, AJ, 124, 1060
960:
961: \bibitem[\protect\citeauthoryear{{Saar} \& {Bookbinder}}{{Saar} \&
962: {Bookbinder}}{2003}]{Saar03}
963: {Saar} S.~H., {Bookbinder} J.~A., 2003, in {Brown} A., {Harper}
964: G.~M.,
965: {Ayres} T.~R., eds, The Future of Cool-Star Astrophysics: 12th Cambridge
966: Workshop on Cool Stars, Stellar Systems, and the Sun (University of Colorado),
967: Vol.~12, {STIS Far UV Studies of Spatial and Temporal
968: Activity Variations in YY Gem}.
969: pp 1020--1023
970:
971: \bibitem[\protect\citeauthoryear{{Schmitt}, {Drake}, {Haisch} \&
972: {Stern}}{{Schmitt} et~al.}{1996}]{Schmitt96}
973: {Schmitt} J.~H.~M.~M., {Drake} J.~J., {Haisch} B.~M., {Stern}
974: R.~A.,
975: 1996, ApJ, 467, 841
976:
977: \bibitem[\protect\citeauthoryear{{Schrijver} \& {Zwaan}}{{Schrijver} \& {Zwaan}}
978: {2000}]{Schrijver00}
979: {Schrijver} C.~J., {Zwaan} C., 2000, Solar and Stellar Magnetic
980: Activity. Cambridge Univ. Press, United Kingdom, p. 263
981:
982: \bibitem[\protect\citeauthoryear{{Siarkowski}}{{Siarkowski}}{1992}]{Siarkowski92}
983: {Siarkowski} M., 1992, MNRAS, 259, 453
984:
985: \bibitem[\protect\citeauthoryear{{Siarkowski}, {Pres}, {Drake}, {White} \&
986: {Singh}}{{Siarkowski} et~al.} {1996}]{Siarkowski96}{Siarkowski}
987: M., {Pres} P., {Drake} S.~A., {White} N.~E., {Singh} K.~P., 1996,
988: ApJ, 473, 470
989:
990: \bibitem[\protect\citeauthoryear{{Stelzer}, {Burwitz}, {Audard}, {G{\"u}del},
991: {Ness}, {Grosso}, {Neuh{\"a}user}, {Schmitt}, {Predehl} \&
992: {Aschenbach}}{{Stelzer} et~al.}{2002}]{Stelzer02}
993: {Stelzer} B., {Burwitz} V., {Audard} M., {G{\"u}del} M.,
994: {Ness} J.-U.,
995: {Grosso} N., {Neuh{\"a}user} R., {Schmitt} J.~H.~M.~M., {Predehl} P.,
996: {Aschenbach} B., 2002, A\&A, 392, 585
997:
998: \bibitem[\protect\citeauthoryear{{Struve} \& {Zebergs}}{{Struve} \&
999: {Zebergs}}{1959}]{Struve59}
1000: {Struve} O., {Zebergs} V., 1959, ApJ, 130, 783
1001:
1002: \bibitem[\protect\citeauthoryear{{Torres} \& {Ribas}}{{Torres} \&
1003: {Ribas}}{2002}]{Torres02} {Torres} Guillermo, {Ribas} Ignasi,
1004: 2002, ApJ, 567, 1140
1005:
1006: \bibitem[\protect\citeauthoryear{{Tsikoudi} \& {Kellett}}{{Tsikoudi} \&
1007: {Kellett}}{2000}]{Tsikoudi00}
1008: {Tsikoudi} V., {Kellett} B.~J., 2000, MNRAS, 319, 1147
1009:
1010: \bibitem[\protect\citeauthoryear{{Uchida} \& {Sakurai}}{{Uchida} \&
1011: {Sakurai}}{1983}]{Uchida83}{Uchida} Y., {Sakurai} T., 1983, in
1012: {Byrne} P.~B., {Rodono} M., eds, IAU Colloq. 71: Activity in Red-Dwarf
1013: Stars Vol.~102 of Astrophysics and Space Science Library, {Interacting
1014: magnetospheres in RS CVn binaries - Coronal heating and flares}.
1015: pp 629--632
1016:
1017: \bibitem[\protect\citeauthoryear{{Uchida} \& {Sakurai}}{{Uchida} \&
1018: {Sakurai}}{1985}]{Uchida85}
1019: {Uchida} Y., {Sakurai} T., 1985, in {Kundu} M.~R., {Holman}
1020: G.~D., eds,
1021: Unstable Current Systems and Plasma Instabilities in Astrophysics Vol.~107 of
1022: IAU Symposium, {Magnetodynamical processes in interacting magnetospheres of
1023: RS CVn binaries}.
1024: pp 281--285
1025:
1026: \bibitem[\protect\citeauthoryear{{van Gent}}{{van Gent}}{1931}]{vanGent31}
1027: {van Gent} H., 1931, BAN.., 6, 99
1028:
1029: \end{thebibliography}
1030:
1031:
1032: \bsp
1033:
1034: \label{lastpage}
1035:
1036: \end{document}
1037: