astro-ph0702512/YS.tex
1: % mn2eguide.tex
2: % v2.1 released 03/05/2002
3: %
4: % Adapted from mnguide.tex
5: % v1.3 released 14th September 1995
6: % v1.2 released 5th September 1994 (M. Reed)
7: % v1.1 released 18th July 1994
8: % v1.0 released 28th January 1994
9:  
10: % The journal style files and macros, with guides on their use, are
11: % available by anonymous FTP on the Internet from the Comprehensive
12: % TeX Archive Network (CTAN) sites ftp.tex.ac.uk and ftp.dante.de.
13: % The files are in the directories
14: % /tex-archive/macros/plain/contrib/mnras and
15: % /tex-archive/macros/latex209/contrib/mnras for the TeX and LaTeX
16: % files respectively.
17: 
18: %\documentclass[useAMS,usenatbib,referee]{mn2e}
19: \documentclass[useAMS,usenatbib]{mn2e}
20: 
21: 
22: \usepackage{epsfig}
23: \renewcommand{\vec}[1]{\mbox{\boldmath $#1$}}
24: 
25: 
26: \title[Polarization of the Crab Nebula]
27: {Polarization of the Crab Nebula with disordered magnetic components}
28: 
29: \author[Y.~Nakamura, S.~Shibata]
30:   {Y.~Nakamura,$^1$\thanks{yuji@ksirius.kj.yamagata-u.ac.jp}
31:   S.~Shibata,$^2$\thanks{shibata@sci.kj.yamagata-u.ac.jp} \\
32:   $^1$Graduate School of Science and Engineering, Yamagata University, Yamagata
33: 990-8560, Japan\\
34:   $^2$Department of Physics, Yamagata University, Yamagata 990-8560, Japan }
35: 
36: \date{Accepted 2007 June 28. Received 2007 June 21; in original form 2007 February 6}
37: 
38: \pagerange{\pageref{firstpage}--\pageref{lastpage}} \pubyear{2007}
39: 
40: \def\LaTeX{L\kern-.36em\raise.3ex\hbox{a}\kern-.15em
41:     T\kern-.1667em\lower.7ex\hbox{E}\kern-.125emX}
42: 
43: \newtheorem{theorem}{Theorem}[section]
44: 
45: \begin{document}
46: \label{firstpage}
47: 
48: \maketitle
49: 
50: \begin{abstract}
51: In this paper, we present an expanding disc model to derive polarization properties of the Crab nebula.
52: The distribution function of the plasma and the energy density of the magnetic field are prescribed 
53: as function of the distance from the pulsar 
54: by using the model by Kennel and Coroniti (1984) with $\sigma = 0.003$, 
55: where $\sigma$ is the ratio of Poynting flux to the kinetic energy flux 
56: in the bulk motion just before the termination shock. 
57: Unlike previous models, we introduce disordered magnetic field, 
58: which is parameterized by the fractional energy density of the disordered component. 
59: Flow dynamics is not solved. The mean field is toroidal.
60: 
61: Averaged polarization degree over the disc is obtained as a function 
62: of inclination angle and fractional energy density of the disordered magnetic field.
63: It is found for the Crab that the disordered component has about $60$ percent of the magnetic field energy. 
64: This value is also supported by the facts 
65: that the disc appears not `lip-shape' but as `rings' in the intensity map as was observed, 
66: and that the highest polarization degree of $\sim 40$ percent is reproduced for rings, 
67: being consistent with the observation.
68: 
69: We suggest that because the disordered field contributes rather pressure than tension, 
70: the pinch force may be over-estimated in previous relativistic magnetohydrodynamic simulations. 
71: Disruption of the toroidal magnetic field with alternating direction, 
72: which is proposed by Lyubarski (2003), may actually takes place. 
73: The relativistic flow speed, 
74: which is indicated by the front-back contrast, 
75: can be detected in asymmetry in distributions of the position angle and depolarization.
76: \end{abstract}
77: 
78: \begin{keywords}
79: pulsars: general -- ISM: individual: Crab nebula -- supernova remnants -- radiation mechanisms: non-thermal -- polarization.
80: \end{keywords}
81: 
82: \section{Introduction}
83: \label{intro}
84: It is well established that the Crab Nebula shines in synchrotron radiation. 
85: However the origin of the high energy particles has been a long-standing problem.
86: The pulsar wind from the central pulsar is certainly the source of the particles and the magnetic fields.
87: We nevertheless do not know what fraction of the wind energy
88: is in magnetic field and in plasma kinetic energy. 
89: We sometimes introduce the $\sigma$-parameter, which is the ratio of
90: the magnetic energy flux to the kinetic one of the wind bulk motion just before the termination shock.
91: 
92: Kennel and Coroniti (KC) (1984a, 1984b) proposed a simple model in which
93: the nebula is expanding hot plasmas after the termination shock
94: of the wind. In their model, the kinetic energy of the wind bulk motion
95: converts into heat with non-thermal components.
96: This picture results in a small value of $\sigma$, i.e., dominance of the kinetic energy, to explain the observed
97: synchrotron luminosity and the presumed expansion velocity of $\sim 2000 \ {\rm km} \; {\rm s}^{-1}$.
98: One finds it difficult to reproduce the kinetic energy dominated wind by means of 
99: the axisymmetric ideal-MHD model (Michel 1969, Bogovalov 1998, Begelman \& Li 1994, Tomimatsu 1994), 
100: although some possibilities for the small $\sigma$ 
101: have been pointed out (Okamoto 2002, Vlahakis 2004). 
102: This is referred to as the $\sigma$-problem.
103: 
104: Because the magnetic axis inclines to the rotation axis,
105: the magnetic
106: neutral sheet in the equator is folded for each rotation to produce
107: a series of current sheets flowing out with the wind.
108: Some authors attempt to solve the $\sigma$-problem by introducing
109: dissipation of the magnetic field in such current sheets
110: (Coroniti 1990, Kirk \& Skjaeraasen 2003).
111: 
112: Lyubarski (2003) has proposed that conversion of the magnetic energy to
113: heat takes place when the current sheets come into the termination shock.
114: With this picture, Poynting energy directly changes to plasma heat and
115: non-thermal particles, so that
116: the pulsar wind should not always be kinetic energy dominant, i.e.,
117: the $\sigma$-problem may not be a problem.
118: 
119: Magnetic field of the wind is essentially toroidal far beyond the light cylinder owing to rotation.
120: The nebula field convected from the wind is also believed to be toroidal,
121: indicated by highly axisymmetric structures in the Crab Nebula.
122: However, if disruption of the numerous current sheets takes place at or after the termination shock, 
123: one may expect that the nebula field is not pure toroidal 
124: but it has strong disordered components.
125: Thus, the disordered magnetic field may be a good indicator for energy conversion 
126: from the toroidal magnetic field to plasmas.
127: To diagnose disruptive processes of the magnetic field, 
128: it must be very much helpful to measure magnitude of the disordered field.
129: 
130: Disordered magnetic field can be detected by image analysis and by polarization analysis. 
131: Shibata et al. (2003; Paper I) calculated images by use of the KC model and compared with the Chandra observations.
132: If the nebula field is pure toroidal, intensity along the major axis of the torus should be reduced, 
133: and the images look like `lips' rather than rings. 
134: The X-ray images, which appear as rings, indicate 
135: that disordered field may be comparable to the mean field.
136: 
137: Polarization observation of the nebula provides the mean field structure and degree of randomness of the field. 
138: The polarization of the Crab Nebula is observed in radio bands (Wilson 1972, Velusamy 1985), 
139: in optical bands (Oort \& Walraven 1956, Schdmidt et al. 1979, Hickson \& van der Bergh 1990, and Michel et al. 1991), 
140: and in X-ray bands (Weisskopf et al. 1978).
141: The X-ray polarization is not spatially-resolved but a mean polarization is obtained to be $\sim 20 \%$. 
142: Almost the same value is obtained also in optical bands (Oort \& Walraven 1956). 
143: 
144: In this paper, we calculate polarization properties of a nebula which expands with relativistic speeds 
145: and with disordered magnetic field as well as mean toroidal fields. 
146: We then compare the results with the Crab observations, 
147: so that the magnitude of the disordered magnetic field is estimated.
148: 
149: \section{Polarization model}
150: \subsection{Polarization from a medium in relativistic motion}
151: \label{uniB}
152: Let us consider synchrotron radiation from a relativistic plasma moving relative to the observer at the velocity \vec{V}.
153: The volume emissivities in terms of the Stokes parameters 
154: in the `flow frame' comoving with the plasma may be given by (Rybicki \& Lightman 1979)  
155: \begin{eqnarray}
156: \left(
157: \begin{array}{c}
158: d I^\prime / d s \\
159: d Q^\prime / d s \\
160: d U^\prime / d s \\
161: d V^\prime / d s
162: \end{array}
163: \right)
164: &=&
165: \left(
166: \begin{array}{c}
167: j^\prime_{tot} \\
168: j^\prime_{pol} \\
169: 0 \\
170: 0
171: \end{array}
172: \right) \nonumber \\
173: &=&
174: \left(
175: \begin{array}{c}
176: \frac{p+7/3}{p+1} ~ \Phi(\omega^\prime,p) ~ (B^\prime \sin{\theta^\prime})^{\frac{p+1}{2}}\\
177: \Phi(\omega^\prime,p) ~ (B^\prime \sin{\theta^\prime})^{\frac{p+1}{2}}\\
178: 0\\
179: 0
180: \end{array}
181: \right), \label{Stok_0}
182: \end{eqnarray}
183: where
184: \begin{eqnarray}
185: \Phi(\omega^\prime,p) = \frac{\sqrt{3} e^3 K}{8 \pi m c^2} \left( \frac{3 e}{m c \omega^\prime} \right)^{\frac{p-1}{2}} \Gamma \left( \frac{p}{4} + \frac{7}{12} \right) \Gamma \left( \frac{p}{4} - \frac{1}{12} \right), 
186: \end{eqnarray}
187: $B^\prime=|\vec{B}^\prime|$ is the magnetic field strength, 
188: $\Gamma$ is the gamma function, 
189: $-e, m$ are the charge and the rest mass of an electron, and $c$ is the speed of light.
190: Here, we have assumed for a given volume element in which 
191: the magnetic field $\vec{B}^\prime$ is uniform, 
192: and the relativistic electrons and positrons have a power-law energy distribution, 
193: $f(\gamma) = K \gamma^{-p}$, where $\gamma$ is the Lorentz factor of the particles; 
194: $p$ and $K$ are functions of position.
195: The pitch angle distribution is assumed to be uniform.
196: In the flow frame, the magnetic field $\vec{B}^\prime$ and the unit vector toward the observer $\vec{n}^\prime$ make the angle $\theta^\prime$, 
197: and the angular frequency of the radiation observed at $\omega$ is Doppler-shifted to $\omega^\prime$.
198: 
199: If the ideal-MHD condition \vec{E} $+$ \vec{\beta} $\times$ \vec{B} = \vec{0} holds, 
200: where \vec{E} and \vec{B} are the electric and magnetic fields in the `observers frame', 
201: and \vec{\beta} $=$ \vec{V} $/c$, 
202: then the external electric field in the flow frame disappears.
203: In this case, we can use the argument by  Bj\"{o}rnson (1982) that 
204: `the observed position angles' $\xi$ 
205: in the flow frame and the observer frame are the same (see Fig.~\ref{flow_obs}).
206: The Stokes parameters for the observer follow 
207: (Blandford \& K\"{o}nigl 1979, Bj\"{o}rnson 1982, Lyutikov et al. 2003)
208: \begin{eqnarray}
209: \left(
210: \begin{array}{c}
211: d I_\omega / d s \\
212: d Q_\omega / d s \\
213: d U_\omega / d s \\
214: d V_\omega / d s
215: \end{array}
216: \right)
217: = 
218: {\mathcal D}^2
219: \left(
220: \begin{array}{c}
221: j^\prime_{tot} \\
222: - \cos{2 (\chi_0 - \xi)} ~ j^\prime_{pol} \\
223: - \sin{2 (\chi_0 - \xi)} ~ j^\prime_{pol} \\
224: 0
225: \end{array}
226: \right) \label{Stok}, 
227: \end{eqnarray}
228: where $\Gamma_f = (1-\beta^2)^{-1/2}$ and ${\mathcal D} = \Gamma^{-1}_f (1-\beta \cos{\alpha})^{-1}$ 
229: are the Lorentz factor and the beam factor of the flow, 
230: respectively, 
231: $\alpha$ is the angle between the direction of the observer \vec{n} 
232: and the flow velocity \vec{V}, $\chi_0$ is the offset caused by the practical position angle which is measured from the north 
233: on the sky for a given observer, 
234: and $\omega^\prime = \omega / {\mathcal D}$.
235: Numerical integration along the line of sight will be done by using grid cells constracted in the observer's frame (see Section 2.2).
236: Quantities in (3) such as ${\mathcal D}$, $\chi_0$, $\xi$ and $\omega^\prime$ are 
237: different in each grid cell because flow velocities are different cell by cell.
238: Suitable transformations are applied for each cell.
239: To each cell, the orthogonal coordinates are constructed by the unit vectors
240: $ \hat{\vec{V}} \equiv \vec{V}/|\vec{V}|$ in the $x$-direction, 
241: $ \vec{n}^\prime \times \hat{\vec{V}}$ in the $y$-direction and 
242: $ \hat{\vec{V}} $ $\times$ ( $ \vec{n}^\prime \times \hat{\vec{V}} $ ) in the $z$-direction, respectively.
243: The observed position angle $\xi$ is measured from the projected $z$-axis on the sky, i.e., 
244: \begin{eqnarray}
245: \tan{\xi} = - \frac{B^\prime_{z}}{B^\prime_{y}} \cos{\alpha^\prime} + \frac{B^\prime_{x}}{B^\prime_{y}} \sin{\alpha^\prime} ,
246: \end{eqnarray}
247: where $B^\prime_x$, $B^\prime_y$, $B^\prime_z$ are the magnetic field components in the flow frame. 
248: The offset angle $\chi_0$ is defined by the equation $\chi_0 \equiv -\cos^{-1}{(\hat{\mathbf Z} \cdot \hat{\mathbf z}^{\prime\prime})}$, 
249: where $\hat{\mathbf z}^{\prime\prime}=-\sin{\alpha} \hat{\mathbf V} + \cos{\alpha} \hat{\mathbf z}$ and $\hat{\mathbf Z}$ 
250: is the unit vector of the north direction on the sky. 
251: In Paper I, we used the transformation of the volume emissivity with the factor $\Gamma_f {\mathcal D}^3$, 
252: and this is not correct. 
253: The factor ${\mathcal D}^2$ should be used. 
254: Fig. 2(bottom) of Paper I may be replaced by Fig.~\ref{resultKC}(a) of the present paper. 
255: However, there is no apparent difference between them; the lip shape persists.
256: 
257: \begin{figure}
258: \begin{center}
259: \includegraphics[clip,width=8.4cm]{figure1a.eps}
260: \includegraphics[clip,width=8.4cm]{figure1b.eps}
261: \end{center}
262: \caption{ \label{flow_obs}
263: (a) and (b) show the flow frame and the observer frame, respectively.
264: The position angle $\xi$ is an invariant.
265: The unit vectors of the flow velocity, the observer's direction and the electric vector of the linearly polarized wave are indicated by $\hat{\vec{V}}$, $\vec{n}$ and $\vec{e}$, respectively.}
266: \end{figure}
267: 
268: \subsection{Construction of polarization map}
269: \label{const}
270: If the nebula is optically thin, it is straightforward to obtain
271: the observed Stokes parameters, simply by integration of (\ref{Stok}) along the line of sight.
272: We use the observer's frame ($X$,$Y$,$Z$) such that 
273: the observer locates at $X=+ \infty$, 
274: the $YZ$ plane defines the sky with the north in $Z$-direction, 
275: and the nebula center (pulsar) is located at the origin.
276: The volume emissivities of (\ref{Stok}) are functions of $X$,$Y$,$Z$ and \vec{n}, 
277: so that maps of the stokes parameters are given by 
278: \begin{eqnarray}
279: \begin{array}{l}
280: I_{\omega} (Y,Z) = \displaystyle{\int^{+\infty}_{-\infty} \frac{dI_{\omega} (X,Y,Z,\vec{n})}{dX} dX},\\ 
281: Q_{\omega} (Y,Z) = \displaystyle{\int^{+\infty}_{-\infty} \frac{dQ_{\omega} (X,Y,Z,\vec{n})}{dX} dX},\\
282: U_{\omega} (Y,Z) = \displaystyle{\int^{+\infty}_{-\infty} \frac{dU_{\omega} (X,Y,Z,\vec{n})}{dX} dX}. 
283: \end{array}\label{INT}
284: \end{eqnarray}
285: We construct grid points on the plane of sky ($Y_i$,$Z_j$) for polarization maps, 
286: and on the line of sight $X_k$ for integration (\ref{INT}) with increment of $\Delta s$. 
287: The Stokes parameters as the whole nebula are obtained by 
288: \begin{eqnarray}
289: \begin{array}{l}
290: I_{t} = \displaystyle \sum_{i} \displaystyle \sum_{j} I(Y_i,Z_j),\\
291: Q_{t} = \displaystyle \sum_{i} \displaystyle \sum_{j} Q(Y_i,Z_j),\\
292: U_{t} = \displaystyle \sum_{i} \displaystyle \sum_{j} U(Y_i,Z_j),
293: \end{array}
294: \end{eqnarray}
295: with which the mean polarization degree $\bar{P}$ and position angle (P.A.), $\bar{\chi}$, 
296: are obtained by 
297: $\bar{P} = \sqrt{{Q_t}^2 + {U_t}^2}/I_t$ and 
298: $\bar{\chi} = (1/2) \tan^{-1} \left( U_t/Q_t \right)$, respectively.
299: 
300: \subsection{Crab Nebula model}
301: \label{model}
302: We start our calculation with a simple disc model with pure toroidal magnetic field (Paper I).
303: The disc represents the post shock flow with an inner radius $R_s$ located at the shock 
304: and with a constant semi-opening angle $\theta_0$. 
305: The flow is based on KC: 
306: the radial flow velocity $V(R)$ is given (KC), 
307: where $R=(X^2+Y^2+Z^2)^{1/2}$. 
308: The post shock flow is characterized 
309: by the $\sigma$ parameter which is the ratio of Poynting flux 
310: to the kinetic energy flux just before the shock. 
311: The nebula flow suffers adiabatic and synchrotron losses.
312: In Paper I, 
313: we calculate the evolution of distribution function. 
314: The parameters $K$ and $p$ in (\ref{Stok_0}) 
315: are determined as functions of $R$ 
316: so as to fit the distribution function obtained 
317: in Paper I at each point in the nebula. 
318: The magnetic field distribution $B_{KC} (R)$ is also given by the KC model.
319: We use following a parameter set: 
320: $\sigma=0.003$, 
321: the wind luminosity $L_w=5 \times 10^{38}$ erg~s$^{-1}$, 
322: the wind Lorentz factor $\gamma_w = 3 \times 10^6$, 
323: the shock distance $R_s = 3 \times 10^{17}$ cm, 
324: the power-law index at the shock $p_s = 3$, 
325: and the thickness of the disc $\theta_0 = \pm 10^\circ$ degree. 
326: The inclination angle of the axis of the disc to the observer is $i = 28^\circ$ (Weisskopf et al. 2000).
327: 
328: \subsection{The case of disordered toroidal field}
329: \label{disB}
330: It has been suggested in Paper I that the nebula field is not 
331: `pure toroidal' but may be dominated by disordered fields. 
332: Scale length of the randomness (turbulent spectrum) is not known; 
333: it can be microscopic or just below the resolution of observation.
334: In this section, let us consider the case of disordered magnetic field. 
335: 
336: In general, the magnetic field can be decomposed to the mean field $\vec{B}^\prime_0$ 
337: and the random field $\vec{B}^\prime_1$. 
338: The degree of randomness may be characterized by 
339: \begin{eqnarray}
340: b= \frac{<{\vec{B}^\prime_1}^2>}{{B^\prime_0}^2+<{\vec{B}^\prime_1}^2>}, 
341: \end{eqnarray}
342: where $< >$ indicates the spatial average in a scale below the observational resolution 
343: and the primes indicate that $b$ is evaluated in the flow frame ($<{\vec{B}^\prime_1}^2>$ is still a function of position). 
344: The synchrotron radiation for such cases is studied by Korchakov \& Syrovat-skii (KS) (1962). 
345: From (3), including relativistic motion of the flow, 
346: we have the observed Stokes parameters: 
347: \begin{eqnarray}
348: \left(
349: \begin{array}{c}
350: d I_\omega / d s \\
351: d Q_\omega / d s \\
352: d U_\omega / d s \\
353: d V_\omega / d s
354: \end{array}
355: \right)
356: = 
357: {\mathcal D}^2
358: \left(
359: \begin{array}{c}
360: j^\prime_{tot} \\
361: - \cos{2 (\chi_0 - \xi_0)} ~ j^\prime_{pol} \\
362: - \sin{2 (\chi_0 - \xi_0)} ~ j^\prime_{pol} \\
363: 0
364: \end{array}
365: \right), \label{Stok_random} 
366: \end{eqnarray}
367: where 
368: \begin{eqnarray}
369: \begin{array}{l}
370: j^\prime_{tot} = \displaystyle{\frac{p+7/3}{p+1} \Phi(\omega^\prime, p)} \\
371: \quad \times \displaystyle{\frac{1}{\Delta s} \int^{s+ \Delta s}_{s} \textstyle{\left( B^\prime \sqrt{1-\left(\vec{n}^\prime \cdot \frac{\vec{B}^\prime}{\displaystyle{B^\prime}}\right)^2} \right)}^{\frac{p+1}{2}} ds^\prime}, \label{j_prime_tot_random}\\
372: \end{array}
373: \end{eqnarray}
374: \begin{eqnarray}
375: \begin{array}{l}
376: j^\prime_{pol} = \displaystyle{\Phi(\omega^\prime, p)} \\
377: \quad \times \displaystyle{\frac{1}{\Delta s} \int^{s+ \Delta s}_{s} \textstyle{\left( B^\prime \sqrt{1-\left(\vec{n}^\prime \cdot \frac{\vec{B}^\prime}{\displaystyle{B^\prime}}\right)^2} \right)}^{\frac{p+1}{2}} \cos{2 \chi} ~ds^\prime}, \label{j_prime_pol_random}\\
378: \end{array}
379: \end{eqnarray}
380: \begin{eqnarray}
381: \begin{array}{l}
382: B^\prime = \displaystyle{\sqrt{ | \vec{B}^\prime |^2} = \sqrt{ {B^\prime_0}^2 + 2 \vec{B}^\prime_{0} \cdot \vec{B}^\prime_{1} + {B^\prime_{1}}^2 } }, 
383: \end{array}
384: \end{eqnarray}
385: \begin{eqnarray}
386: \begin{array}{l}
387: \cos{2 \chi} = \displaystyle{2 \left( \frac{ \vec{B}^\prime_{0} \times \vec{n}^\prime}{|\vec{B}^\prime_0 \times \vec{n}^\prime|} \cdot \frac{ \vec{B}^\prime \times \vec{n}^\prime}{|\vec{B}^\prime \times \vec{n}^\prime|} \right)^2 -1 } \label{cos2chi}.
388: \end{array}
389: \end{eqnarray}
390: The volume emissivities (\ref{Stok_random}) are obtained for each grid cell 
391: and are integrated numerically according to (\ref{INT}). 
392: Given by a nebula model, 
393: macroscopic quantities such as $\vec{V}$ and $\vec{B}_0$ 
394: are assumed to be constant in a given grid cell, 
395: but gradually change cell by cell. 
396: In each cell, averaging for the random field (\ref{j_prime_tot_random}) and (\ref{j_prime_pol_random}) is done by Monte Carlo method. 
397: KS use the special coordinate in which $U$ vanishes. 
398: By using this coordinate, only $Q$ remains, 
399: and Monte Carlo integration can be only one time; 
400: the factor $\cos{2 \chi}$ appears in (\ref{j_prime_pol_random}), 
401: but does not $\sin{2 \chi}$, 
402: where $\chi$ is defined by (\ref{cos2chi}) and represents the angle 
403: between the polarization E-vector and the projected magnetic field. 
404: The value of $\chi$ varies place by place within the cells. 
405: Redistribution of the Stokes parameters for the observers coordinate can be done by (\ref{Stok_random}) 
406: by using the angle $\xi_0$ defined by the projected mean field: 
407: \begin{eqnarray}
408: \tan{\xi_0} = - \frac{B^\prime_{0z}}{B^\prime_{0y}} \cos{\alpha^\prime} + \frac{B^\prime_{0x}}{B^\prime_{0y}} \sin{\alpha^\prime}. 
409: \end{eqnarray}
410: 
411: As stated above, 
412: integration along the line of sight has been done in use of Monte Carlo method with randomly distributed $\vec{B}^\prime_1$'s. 
413: In the present calculation, 
414: we assume for comparison purposes that the energy density 
415: of the nebula field is equal to the KC value $B^2_{KC}$, 
416: i.e., 
417: ${B^\prime_0}^2 + <{\vec{B}^\prime_1}^2> = B^2_{KC}/\Gamma^2_f$, 
418: and that the plasma's distribution functions are the same as Paper I.
419: It is obvious that the disordered fields change flow dynamics: 
420: (i) the disordered fields contribute to the magnetic pressure rather than magnetic tension, 
421: (ii) formation of the disordered fields must be associated by heating, 
422: and (iii) the distribution function must be changed. 
423: These issues are postponed to subsequent papers.
424: 
425: Apart from the disordered field, it is also suggested by 
426: some authors (Hester et al. 2002, Mori et al. 2004, Ng \& Romani 2004, and Paper I) 
427: that if the front-back (north-west to south-east) 
428: intensity contrast is due to Doppler boost, 
429: the flow velocity is much faster than the KC flow ($\sim 2000 \ {\rm km} \; {\rm s}^{-1}$) and 
430: will be $\sim 0.2 c$. 
431: Therefore, we also calculate the case with much faster flows. 
432: 
433: In summary, we calculate three cases; 
434: Case KC, in which plasma properties follows the KC model (see Section 2.3 in detail); 
435: Case D, in which we include the disordered field with $b$ (the degree of randomness). 
436: For comparison purposes the total magnetic energy density and 
437: flow speed are left unchanged; 
438: Case DR, in which, in addition to Case D 
439: we change the flow velocity to be $0.2 c$ through out the whole nebula 
440: by hand (we ignore the flow dynamics). 
441: 
442: 
443: \section{Results}
444: \subsection{Case of Pure Toroidal Field}
445: \label{RuniB}
446: \begin{figure}
447: \begin{center}
448:  \begin{minipage}{0.48\linewidth}
449:    \includegraphics[angle=-90,width=\linewidth]{figure2a.eps}
450:  \end{minipage}
451:  \begin{minipage}{0.48\linewidth}
452:    \includegraphics[angle=-90,width=\linewidth]{figure2b.eps}
453:  \end{minipage}
454: \\[5mm]
455:  \begin{minipage}{0.48\linewidth}
456:    \includegraphics[angle=-90,width=\linewidth]{figure2c.eps}
457:  \end{minipage}
458:  \begin{minipage}{0.48\linewidth}
459:    \includegraphics[width=\linewidth]{figure2d.eps}
460:  \end{minipage}
461: \end{center}
462: \caption{ \label{resultKC}
463: Polarization properties of the pulsar disc nebula for Case KC, 
464: observed at $5.2$ keV.
465: (a)Intensity map. 
466: The unit length is a shock radius and the unit of the intensity is 0.016 erg~s$^{-1}$~cm$^{-2}$~str$^{-1}$~eV$^{-1}$. 
467: (b)Polarization degree map 
468: drawn only for the regions where the intensity is strong enough, regarding calculation errors.
469: (c)P.A. map (E-vector). 
470: (d)Polarization degree as a function 
471: of the azimuthal angle along an ellipse with the 
472: semi-major axis of $2.5 R_s$ and semi-minor axis of $1.5 R_s$.
473: The labels `SW', `NE' and `SW' below the horizontal axis 
474: indicate south-west, north-east and south-west, respectively.
475: }
476: \end{figure}
477: Fig.~\ref{resultKC} shows the results of Case KC.  
478: The observation energy is assumed to be $5.2$ keV.
479: The intensity map (Fig.~\ref{resultKC}(a)) 
480: shows a lip-shaped nebula, 
481: as was pointed out in Paper I.
482: On the major axis (NE-SW) of the ring (perpendicular to the rotation axis of the pulsar) and its vicinity, 
483: the pitch angles of the particles radiating toward the observer are small, 
484: and as a result synchrotron emissivity is decreased. 
485: The dimness is cased not only by decrease of single-particle emissivity 
486: but also by decrease of number of the particles emitting to the observing bands. 
487: Fig.~\ref{resultKC}(b) shows the polarization degree. 
488: In Fig.~\ref{resultKC}(d), the polarization degree 
489: is plotted as an azimuthal function along the ring. 
490: On the minor axis (rotational axis of the pulsar), 
491: the polarization degree reaches the maximum value of $\sim  75\% $.
492: There is depolarized regions along the major axis, 
493: where the polarization degree decreases to $\sim 55\%$. 
494: 
495: In Case KC, 
496: the mean polarization degree as the whole nebula is $52.3 \%$, 
497: and P.A. is found to be $132^\circ$ at $5.2$ keV. 
498: This polarization degree is much larger than the observed value ($\sim 20 \%$) in X-ray (Weisskopf et al. 1978).
499: Although we calculate also for optical bands, 
500: the results are hardly changed so that they are not shown in figures.
501: Since the mean polarization in optical is $\sim 19 \%$ (Oort and Walraven 1956), 
502: the calculated polarization degree in optical is again much higher than the observed. 
503: \begin{figure}
504: \begin{center}
505: \includegraphics[clip,width=8.4cm]{figure3.eps} 
506: \end{center}
507: \caption{ \label{Pofbi}
508: Polarization of the entire nebula as functions 
509: of the inclination angle $i$ between the axis of the disc and the line of sight for $b=0.0$, $0.2$, $0.4$, $0.6$ and $0.8$.
510: The empirical curves $\bar{P} = 75.8 (1-b) (\cos{i})^{2.84}$ are overlaid.
511: }
512: \end{figure}
513: 
514: \subsection{Effects of Disordered Field and Relativistic Motion}
515: \label{RdisB}
516: In Case D, 
517: we change the parameter $b$ to find the value which reproduces the observed polarization degree.
518: Fig.~\ref{Pofbi} shows how the mean polarization behaves 
519: as a function of $b$ and the inclination angle $i$, 
520: the angle of the symmetry axis of the nebula from \vec{n} (the line of sight). 
521: The result can be fitted by an empirical relation, 
522: $\bar{P} = 75.8 \times (1-b) \times (\cos{i})^{2.84}$. 
523: For the Crab Nebula, the inclination angle is known to be $i = 28^\circ$ (Weisskopf et al. 2000), 
524: and therefore the observed polarization degree can be reproduced if $b=0.6$. 
525: 
526: Polarization properties for Case D with $b=0.6$ are summarized 
527: in Fig.~\ref{resultD} at the observation energy of $5.2$ keV. 
528: As has been mentioned in Section 1, 
529: almost the same polarization degrees are observed both in X-ray and optical bands. 
530: The calculated polarization properties for the optical bands are hardly different from those for the $5.2$ keV band.
531: This is thus consistent with the observation. 
532: The polarization degree takes the maximum value of $\sim 40 \%$ 
533: on the northwest side and the southeast side of the ring.
534: This also agrees with the observation in optical bands 
535: (Schmidt et al. 1979, Hickson \& van der Bergh 1990, Michel et al. 1991). 
536: 
537: Case D is favored by at least two observational facts 
538: that (1) the intensity map appears in a ring (not lip-shaped), 
539: and that (2) the polarization degree is $\sim 40 \%$ 
540: at the highest and $\sim 20 \%$ on average. 
541: This indicates that the disordered magnetic component is $60 \%$ of the total.
542:   
543: The front-back contrast in Case D is smaller 
544: than the observed value $3.4$ (Mori et al. 2004) 
545: because the model follows the KC flow. 
546: In Case DR, 
547: the flow velocity is assumed to be $0.2 c$ in the whole nebula, 
548: and thereby front-back contrast becomes $3.1$ (Fig.~\ref{resultDR}).
549: The average polarization of Case DR is increased only by $0.14 \%$ as compared with Case D.
550: This increase is caused by the high polarization degree 
551: at the north-west region weighted by Doppler boosting. 
552: Relativistic effects is rather prominent in distribution of the depolarization. 
553: When the nebula flow is non-relativistic as seen in Case KC and Case D, 
554: the depolarization takes place along the major axis symmetrically, 
555: but in Case DR the depolarized regions shift slightly to the northwest side 
556: which expands toward us (Fig.~\ref{resultDR}(b) and (d)). 
557: The electric field vectors of the synchrotron radiation shown in Fig.~\ref{resultDR}(c) is no longer vertical 
558: to the projected magnetic field on the sky as the flow becomes relativistic. 
559: These relativistic effect is difficult to see in the complex optical polarization maps. 
560: However, 
561: it might be verified if spatially-resolved polarization in X-ray bands is performed 
562: because the ring structure has a better symmetry in X-ray, 
563: and obscuring by filaments is less. 
564: After all, 
565: the relativistic effect does not change the estimate of the disordered magnetic field. 
566: 
567: We reproduced the maps of Stokes parameters, 
568: Q-map and U-map in Fig.~\ref{resultQU}, 
569: which are compared with the observation (Michel et al. 1991). 
570: Because interaction of the filaments and the nebula flow changes 
571: the direction of the magnetic field, 
572: comparison may be limited within the area of $1^\prime$ around the pulsar. 
573: The shape of hourglass (Michel et al. 1991) is well reproduced in our Q map. 
574: It follows from this fact that the average magnetic field is toroidal. 
575: Because depolarization occurs along the major axis on which U has positive value, 
576: the absolute value of U is small (see Fig.~\ref{resultQU}). 
577: Therefore, U-map is easy to be influenced by disturbances, 
578: and therefore comparison of the observations to the model is difficult to make. 
579: \begin{figure}
580: \begin{center}
581:  \begin{minipage}{0.48\linewidth}
582:    \includegraphics[angle=-90,width=\linewidth]{figure4a.eps}
583:  \end{minipage}
584:  \begin{minipage}{0.48\linewidth}
585:    \includegraphics[angle=-90,width=\linewidth]{figure4b.eps}
586:  \end{minipage}
587: \\[5mm]
588:  \begin{minipage}{0.48\linewidth}
589:    \includegraphics[angle=-90,width=\linewidth]{figure4c.eps}
590:  \end{minipage}
591:  \begin{minipage}{0.48\linewidth}
592:    \includegraphics[width=\linewidth]{figure4d.eps}
593:  \end{minipage}
594: \end{center}
595: \caption{ \label{resultD}
596: Same as Figure \ref{resultKC} but for Case D with $b=0.6$.
597: }
598: \end{figure}
599: \begin{figure}
600: \begin{center}
601:  \begin{minipage}{0.48\linewidth}
602:    \includegraphics[angle=-90,width=\linewidth]{figure5a.eps}
603:  \end{minipage}
604:  \begin{minipage}{0.48\linewidth}
605:    \includegraphics[angle=-90,width=\linewidth]{figure5b.eps}
606:  \end{minipage}
607: \\[5mm]
608:  \begin{minipage}{0.48\linewidth}
609:    \includegraphics[angle=-90,width=\linewidth]{figure5c.eps}
610:  \end{minipage}
611:  \begin{minipage}{0.48\linewidth}
612:    \includegraphics[width=\linewidth]{figure5d.eps}
613:  \end{minipage}
614: \end{center}
615: \caption{ \label{resultDR}
616: Same as Figure \ref{resultKC} but for Case DR with $b=0.6$, $v_r=0.2c$.
617: (d) shows the comparison between the polarization degree of Case D (solid line) 
618: and that of Case DR (dotted line).
619: }
620: \end{figure}
621:  
622: \section{Summary and Discussion}
623: \label{discus}
624: We calculate the synchrotron polarization properties from a relativistically expanding disk 
625: which has both the toroidal mean field component and the disordered component. 
626: We compared our results with observations of the Crab Nebula. 
627: The distribution function of plasma particles 
628: and the total magnetic energy density 
629: $(B^2_0+B^2_1) / 8 \pi$ in the disk are given as a function of the distance $R$, 
630: according the KC flow with $\sigma = 0.003$. 
631: The best fit model reproducing the observed polarization 
632: indicates that the disordered magnetic field contributes $60 \%$ 
633: of the total magnetic energy.
634: This result is also supported by the fact that 
635: if the magnetic field is pure toroidal, 
636: then the nebula image becomes lip-shaped, 
637: while if there is such a disordered magnetic field, 
638: it becomes ring-shaped as observed. 
639: We also attempt to reproduce the front-back contrast of the ring by changing the flow velocity. 
640: The estimate of disordered field is not changed even if a high speed flow of $0.2c$ is introduced. 
641: It is shown that relativistic motion of the expanding nebula can be detected as asymmetry in P.A. and in distribution of depolarized regions. 
642: 
643: Nothing is known about wavelengths of the disordered magnetic field according to our study.
644: It is pointed out that series of magnetic neutral sheets 
645: in the wind can be collapsed at the shock (Lyubarsky 2003). 
646: If it is so, the disordered magnetic field may be formed 
647: in the process of magnetic reconnection. 
648: The scale lengths of the disordered field is thought of 
649: as $c/\Omega \approx 1.6 \times 10^8$ or less. 
650: 
651: Recently, relativistic magneto-hydrodynamic (RMHD) simulations
652: are done for the flow after the termination shock of the pulsar wind, 
653: and formation mechanism of the disk and the jets is discussed 
654: (Komissarov \& Lyubarsky 2003, Del Zanna et al. 2004).
655: Because of the axisymmetry, the magnetic field is assumed to be pure toroidal in those simulations.
656: Therefore, even though vortex is formed in the meridional plane, 
657: it does not cause significant depolarization (Bucciantini et al. 2005, Del Zanna et al. 2006). 
658: 
659: \begin{figure}
660: \begin{center}
661:  \begin{minipage}{0.48\linewidth}
662:    \includegraphics[width=\linewidth]{figure6a.eps}
663:  \end{minipage}
664:  \begin{minipage}{0.48\linewidth}
665:    \includegraphics[width=\linewidth]{figure6b.eps}
666:  \end{minipage}
667: \\[5mm]
668:  \begin{minipage}{0.48\linewidth}
669:    \includegraphics[angle=-90,width=\linewidth]{figure6c.eps}
670:  \end{minipage}
671:  \begin{minipage}{0.48\linewidth}
672:    \includegraphics[angle=-90,width=\linewidth]{figure6d.eps}
673:  \end{minipage}
674: \end{center}
675: \caption{ \label{resultQU}
676: The comparison between the $Q$, $U$ maps observed in optical bands (Michel et al. 1991) 
677: and those of Case DR ($b=0.6$, $\beta=0.2$). 
678: The upper row is an observation, and the lower is the model.
679: The $Q$ map (left) is N/S polarized light minus E/W polarized light, 
680: the $U$ map (right) is NW/SE polarized light 
681: minus NE/SW polarized light.
682: }
683: \end{figure}
684: 
685: It is notable that if the disordered field dominates, 
686: then the magnetic field contributes more as pressure than tension in the flow dynamics, 
687: as compared with the case of pure toroidal field, 
688: which is assumed in the previous RMHD simulations. 
689: As a result, magnetic pinch effect may be overestimated 
690: in the previous simulations. 
691: If the disordered magnetic field is taken into account, 
692: the jets may be weaken, 
693: and magnetic pressure may cause acceleration of the disk flow.
694: 
695: Hickson (1990) finds highest polarization of $\sim 60 \%$ 
696: in a region $1^\prime.8$ apart from the pulsar in a south-south-east direction 
697: and the eastern bay. 
698: For these regions, very ordered magnetic field is expected. 
699: However, polarization degrees of $35 \%$ for the wisp1, $23 \%$ for the wisp3 (Hickson 1990), 
700: and $\sim 30 \%$ of the wisp3 (Scargle 1971) seem to be consistent with $b \sim 0.6$. 
701: 
702: For a future possibility, 
703: we simulated change of polarization for the entire Crab Nebula 
704: through a lunar occultation. 
705: Fig.~\ref{enpei1} shows the results when the nebula is 
706: gradually hidden, 
707: while Fig.~\ref{enpei2} shows a recovery phase. 
708: The direction of the occultation can be seen in the images of the figures.  
709: In the disappearance phase (Fig.~\ref{enpei1}), 
710: the total polarization degree decreases at first 
711: because of hiding highly polarized parts, 
712: and then it is minimized and increases. 
713: The late phase will not be observed 
714: because the south-east part of the ring is dim, 
715: and the radiation from the jet will dominate.  
716: In the reappearance phase, 
717: the polarization degree start with a very high state 
718: $\sim 40 \%$ and gradually recovers. 
719: There will be no significant change of position angle. 
720: 
721: \begin{figure}
722: \begin{center}
723: \includegraphics[clip,width=8cm]{figure7a.eps}\\
724: \includegraphics[clip,width=4cm]{figure7b.eps}
725: \includegraphics[clip,width=4cm]{figure7c.eps}
726: \end{center}
727: \caption{ \label{enpei1}
728: Polarization at a lunar occultation of the Crab Nebula (disappearance phase). 
729: We use Case DR, $b=0.6$, $\beta=0.2$. 
730: (a)Appearance of luna occultation from the northwest to the southeast along symmetrical axis of the nebula. 
731: (b)Change of averaged polarization degree. 
732: (c)Change of P.A. 
733: The labels 0$\sim$7 of the horizontal axis of panel (b) and (c) 
734: correspond to eight figures from on the upper left to lower right of sub-panels in (a). 
735: }
736: \end{figure}
737: 
738: \begin{figure}
739: \begin{center}
740: \includegraphics[clip,width=8cm]{figure8a.eps}\\
741: \includegraphics[clip,width=4cm]{figure8b.eps}
742: \includegraphics[clip,width=4cm]{figure8c.eps}
743: \end{center}
744: \caption{ \label{enpei2}
745: The same as Fig.~\ref{enpei1}, but for the appearance phase. 
746: The labels 1$\sim$8 of the horizontal axis of panel (b) and (c) 
747: correspond to eight figures from on the upper left to lower right of sub-panels in (a). 
748: }
749: \end{figure}
750: 
751: \section*{Acknowledgments}
752: We would like to thank Dr. S. Gunji, Dr. H. Sakurai, Dr. K. Mori, and Dr. K. Ioka for fruitful discussions.
753: This work was supported Grant-in-Aid for Scientific Research 
754: from the Ministry of Education, Culture, Sports, Science and Technology of Japan (15540227). 
755:  
756: \begin{thebibliography}{}
757: \bibitem[\protect\citeauthoryear{Bj\"{o}rnsson}{1982}]{Bj}
758:    Bj\"{o}rnsson, C.-I., 1982, ApJ, 260, 855
759: \bibitem[\protect\citeauthoryear{Blandford \& K\"{o}nigl}{1979}]{BK}
760:    Blandford, R.D., \& K\"{o}nigl, A., 1979, ApJ, 232, 34
761: \bibitem[\protect\citeauthoryear{Begelman \& Li}{1994}]{BLi}
762:    Begelman, M. C., \& Li, Z.-Y., 1994, ApJ, 426, 269
763: \bibitem[\protect\citeauthoryear{Bogovalov}{1998}]{Bog}
764:    Bogovalov S. V., 1998, Astronomy Letters, 24, 321-331
765: \bibitem[\protect\citeauthoryear{Bucciantini, Del Zanna, Amato \& Volpi}{2005}]{BDAV} 
766:    Bucciantini, N., Del Zanna, L., Amato, E., \& Volpi, D., 2005, A \& A, 443, 519B 
767: \bibitem[\protect\citeauthoryear{Coroniti}{1990}]{Cor}
768:    Coroniti, F.V., 1990, ApJ, 349, 538
769: \bibitem[\protect\citeauthoryear{Del Zanna, Amato \& Bucciantini}{2004}]{DAB}
770:    Del Zanna, L., Amato, E., \& Bucciantini, N., 2004, A \& A, 421, 1063
771: \bibitem[\protect\citeauthoryear{Del Zanna, Volpi, Amato \& Bucciantini}{2006}]{DVAB}
772:    Del Zanna, L., Volpi, D., Amato, E., \& Bucciantini, N., 2006, A \& A, 453, 621D
773: \bibitem[\protect\citeauthoryear{Hester, Mori, Burrows, Gallagher, Graham, Halverson, Kader, Michel \& Scowen}{2002}]{He}
774:    Hester, J. J., Mori, K., Burrows, D., Gallagher, J. S., Graham, J. R., Halverson, M., Kader, A., Michel, F. C., Scowen, P., 2002, ApJ, 577, 49
775: \bibitem[\protect\citeauthoryear{Hickson \& van der Bergh}{1990}]{HvdB}
776:    Hickson, P., \& van der Bergh, S., 1990, ApJ, 365, 224
777: \bibitem[\protect\citeauthoryear{Kennel \& Coroniti}{1984}]{KCa}
778:    Kennel, C.F., \& Coroniti, F.V., 1984a, ApJ, 283, 694
779: \bibitem[\protect\citeauthoryear{Kennel \& Coroniti}{1984}]{KCb}
780:    Kennel, C.F., \& Coroniti, F.V., 1984b, ApJ, 283, 710
781: \bibitem[\protect\citeauthoryear{Kirk \& Skjaeraasen}{2003}]{KiSk}
782:    Kirk J.G., \& Skjaeraasen O., 2003, ApJ, 591, 366
783: \bibitem[\protect\citeauthoryear{Komissarov \& Lyubarsky}{2003}]{KL1}
784:    Komissarov, S.S., \& Lyubarsky, Y.E., 2003, MNRAS, 344, L93
785: \bibitem[\protect\citeauthoryear{Korchakov \& Syrovat-skii}{1962}]{KS} 
786:    Korchakov, A.A., \& Syrovat-skii, S.I., 1962, SvA, 5, 678
787: \bibitem[\protect\citeauthoryear{Lyutikov, Pariev \& Blandford}{2003}]{LPB}
788:    Lyutikov, M., Pariev, V.I., \& Blandford, R.D., 2003, ApJ, 597, 998
789: \bibitem[\protect\citeauthoryear{Lyuvarsky}{2003}]{Lyu}
790:    Lyuvarsky, Y. E., 2003, MNRAS, 345, 153-160
791: \bibitem[\protect\citeauthoryear{Michel}{1969}]{Mic}
792:    Michel, F. C., 1969, ApJ, 158, 727
793: \bibitem[\protect\citeauthoryear{Michel, Scowen, Dufour \& Hester}{1991}]{MSDH}
794:    Michel, F.C., Scowen, P.A., Dufour, R.J., \& Hester, J.J., 1991, ApJ, 368, 463
795: \bibitem[\protect\citeauthoryear{Mori, Burrows, Hester, Pavlov, Shibata \& Tsunemi}{2004}]{Mo}
796:    Mori, K., Burrows, D. N., Hester, J. J., Pavlov, G. G., Shibata, S., \& Tsunemi, H., 2004, ApJ, 609, 186
797: \bibitem[\protect\citeauthoryear{Ng \& Romani}{2004}]{NgR}
798:    Ng, C. -Y., \& Romani, R. W., 2004, ApJ, 601, 479
799: \bibitem[\protect\citeauthoryear{Okamoto}{2002}]{Ok}
800:    Okamoto, I., 2002, ApJ, 573, L31
801: \bibitem[\protect\citeauthoryear{Oort \& Walraven}{1956}]{OW}
802:    Oort, J. H., \& Walraven, T., 1956, Bull. Astr. Inst. Netherlands, 12, 285
803: \bibitem[\protect\citeauthoryear{Rybicki \& Lightman}{1979}]{RL}
804:    Rybicki, G.B., \& Lightman, A.P., 1979, Radiative Processes in Astrophysics, Wiley, New York
805: \bibitem[\protect\citeauthoryear{Scargle}{1971}]{Sc}
806:    Scargle, J.D., 1971, Nature Phys. Sci., 230, 37
807: \bibitem[\protect\citeauthoryear{Schmidt, Angel \& Beaver}{1979}]{SAB}
808:    Schmidt, G.D., Angel, J.R.P., \& Beaver, E.A., 1979, ApJ, 227, 106
809: \bibitem[\protect\citeauthoryear{Shibata, Tomatsuri, Shimanuki, Saito \& Mori}{2003}]{STSSM}
810:    Shibata, S., Tomatsuri, H., Shimanuki, M., Saito, K., \& Mori, K., 2003, MNRAS, 346, 841
811: \bibitem[\protect\citeauthoryear{Tomimatsu}{1994}]{Tomi}
812:    Tomimatsu A., 1994, PASJ, 46 123
813: \bibitem[\protect\citeauthoryear{Velusamy}{1985}]{Ve}
814:    Velusamy, T., 1985, MNRAS, 212, 359
815: \bibitem[\protect\citeauthoryear{Vlahakis}{2004}]{Vl}
816:    Vlahakis, N., 2004, ApJ, 600, 324
817: \bibitem[\protect\citeauthoryear{Weisskopf, Silver, Kestenbaum, Long \& Novick}{1978}]{We1}
818:    Weisskopf, M. C., Kestenbaum, H. L., Long, K. S., Novick, R., Silver, E. H., 1978, ApJ, 220, 117
819: \bibitem[\protect\citeauthoryear{Weisskopf, Hester, Tennant, Elsner, Schulz, Marshall, Karovska, Nichols, Swartz, Kolodziejczak \& O'Dell}{2000}]{We2}
820:    Weisskopf, M. C., Hester, J. J., Tennant, A. F., Elsner, R. F., Schulz, N. S., Marshall, H. L., Karovska, M., Nichols, J. S., Swartz, D. A., Kolodziejczak, J. J., \& O'Dell, S. L., 2000, ApJ, 536, 81
821: \bibitem[\protect\citeauthoryear{Wilson}{1972}]{Wil}
822:    Wilson, A. S., 1972, MNRAS, 157, 229
823: \end{thebibliography}
824: 
825: 
826: \label{lastpage}
827: \end{document}