1: \documentclass[12pt,preprint]{aastex}
2: \newcommand{\vdag}{(v)^\dagger}
3: \newcommand{\myemail}{skywalker@galaxy.far.far.away}
4:
5: \shorttitle{Relativistic reconnection} \shortauthors{Karlick\'y}
6: \begin{document}
7:
8: \title{Separation of accelerated electrons and positrons in the
9: relativistic reconnection}
10:
11: \author{Marian Karlick\'y}
12: \affil{Astronomical Institute of the Academy of Sciences of the Czech Republic,
13: CZ-25165 Ond\v{r}ejov, Czech Republic}
14: \email{karlicky@asu.cas.cz}
15:
16:
17: \begin{abstract}
18: We study an acceleration of electrons and positrons in the relativistic
19: magnetic field reconnection using a 2.5-D particle-in-cell electromagnetic
20: relativistic code. We consider the model with two current sheets and periodic
21: boundary conditions. The electrons and positrons are very effectively
22: accelerated during the tearing and coalescence processes of the reconnection.
23: We found that near the X-points of the reconnection the positions of electrons
24: and positrons differ. This separation process is in agreement with those
25: studied in the previous papers analytically or by test particle simulations. We
26: expect that in dependence on the magnetic field connectivity this local
27: separation can lead to global spatial separation of the accelerated electrons
28: and positrons. A similar simulation in the electron-proton plasma with the
29: proton-electron mass ratio $m_i/m_e$ = 16 is made.
30: \end{abstract}
31:
32:
33: \keywords{Acceleration of particles -- Plasmas - Relativity}
34:
35:
36:
37:
38:
39: \section{INTRODUCTION}
40:
41: Magnetic reconnection is the key process in conversion of the magnetic field
42: energy into particle kinetic energy. It is well accepted that it plays a
43: crucial role in the Earth's magnetotail, solar flares and accretion discs
44: (Priest \& Forbes 2000, Drake et al. 2005, Pritchett 2006). The relativistic
45: reconnection in electron-positron plasmas is proposed for high energy
46: astrophysical phenomena including the jets from active galactic nuclei (Lesch
47: \& Birk 1998, Larrabee et al. 2003, Wardle et al. 1998), pulsar winds (Coroniti
48: 1990, Michel 1994, Lyubarsky \& Kirk 2001) and models of gamma-ray bursts
49: (Drenkhahn 2002a,b). The relativistic reconnection and particle acceleration in
50: pair plasmas was studied numerically for the first time by Zenitani \& Hoshino
51: (2001, 2005). The effectiveness of such an acceleration and corresponding
52: synchrotron spectra has been computed in detail in the paper by Jaroschek et
53: al. (2004a,b). Bessho \& Bhattacharjee (2005) have shown that this fast
54: reconnection is caused by the off-diagonal components of the pressure tensor.
55:
56: Recently, the RHESSI observations of the solar flare on 2002 July 23 have
57: revealed a separation of the gamma-ray source from any of those observed in the
58: hard X-ray emission. It has been interpreted as a spatial separation of
59: energetic electrons and protons (Share et al. 2003). Based on the analytical
60: and test particle approach Zharkova \& Gordovskyy (2004) have explained this
61: separation by the asymmetry in acceleration of electrons and protons in the
62: reconnecting non-neutral current sheet, see also the papers by Martens \& Young
63: (1990), Zhu \& Parks (1993), Litvinenko (1996).
64:
65: This separation acceleration can be even more distinct in pair plasmas due to
66: the same mass of electrons and positrons. Therefore in this paper using the
67: particle-in-cell modelling we study this process in the electron-positron
68: plasma in detail.
69:
70:
71:
72: \section{MODEL}
73:
74: We used a 2.5-D (2D3V -- 2 spatial and 3 velocity components) fully
75: relativistic electromagnetic particle-in-cell code (Saito \& Sakai 2004). The
76: system size is $L_x \times L_y$ = 2000$\Delta$ $\times$ 600$\Delta$ = 200$d_e$
77: $\times$ 60$d_{e}$, where $\Delta$ (=1) is a grid size, $d_e$ = $c/\omega_{pe}$
78: is the electron inertial length, $c$ is the speed of light and $\omega_{pe}$ is
79: the plasma frequency.
80:
81: Two 2-D current sheets with the guiding magnetic field B$_z$ are initiated
82: along the lines $y$ = 150$\Delta$ and $y$ = 450$\Delta$. The periodic boundary
83: conditions are used. The half-width of both the current sheets is 10$\Delta$ =
84: $d_e$. The initial magnetic field is (see also Karlick\'y \& B\'arta 2007)
85: \begin{eqnarray}
86: {\bf B} \equiv ({B_x, B_y, B_z}),\nonumber \\
87: B_x = -B_0~ \rm{for~y} < 140 \Delta ,\nonumber \\ B_x = (y-150) B_0/10~{\rm
88: for}~ 140 \Delta \leq~{\rm y} \leq~160 \Delta, \nonumber \\ B_x = B_0~
89: \rm{for}~ 160
90: \Delta < y < 440 \Delta , \nonumber \\
91: B_x = -(y-450) B_0/10~{\rm for}~ 440 \Delta \leq~{\rm y} \leq~460 \Delta, \nonumber \\
92: B_x = - B_0 ~ \rm{for~y} > 460 \Delta ,\nonumber \\
93: B_y = 0, B_z =B_0. \nonumber
94: \end{eqnarray}
95:
96:
97: We consider the electron-positron plasma. In each numerical cell located out of
98: the current sheet we initiated n$_0$ = 60 electrons and n$_0$ = 60 positrons.
99: In this region out of the current sheet we define the plasma frequency for the
100: time unit $\omega_{pe}^{-1}$. The time step in computations is $\omega_{pe}
101: \Delta t$ = 0.05. The total amount of particles in the model is 172 millions.
102: The initial number density is enhanced in the current sheets just to keep the
103: pressure equilibrium in the current sheet. The particle distribution is taken
104: as by Zenitani \& Hoshino (2001) in the form of f $\sim \exp
105: {-m[u_x^2+u_y^2+(u_z-U)^2]/2T}$, where the velocity $u$ is related to the
106: particle velocity $v$ as $u = v \gamma = v/[1-(v/c)^2]^{1/2}$, $U$ is the drift
107: velocity, $m$ is the electron rest mass, $T$ is the plasma temperature and $c$
108: is the speed of light. The mean initial thermal energy of electrons and
109: positrons is taken the same as 0.45 $mc^2$. We neglect any collisions, pair
110: production, and pair annihilation of pair plasmas.
111:
112: Due to our interest about the reconnection processes in the relativistic plasma
113: with high magnetic field we consider cases with low-$\beta$ plasmas. The plasma
114: beta parameter and the ratio of the electron-cyclotron and electron-plasma
115: frequencies in the region out of the current sheets are chosen as $\beta$=0.11,
116: $\omega_{ce}/\omega_{pe}$ = 4 (Case I) and $\beta$=0.5,
117: $\omega_{ce}/\omega_{pe}$ = 1.9 (Case II). For comparison one run was made for
118: the parameters as in Case I, but without the guiding magnetic field, i.e. $B_z
119: = 0$.
120:
121: Furthermore the same processes are modelled in the electron-proton plasma (Case
122: III) with the proton-electron mass ratio $m_i/m_e$ = 16. The proton and
123: electron temperature is taken the same $T_i=T_e$. The parameters are $\beta$ =
124: 0.11 and $\omega_{ce}/\omega_{pe}$ = 4. The mean initial thermal energy is 0.45
125: $mc^2$.
126:
127: All computations were performed on the parallel computer OCAS (Ond\v{r}ejov
128: Cluster for Astrophysical Simulations), see http://wave.asu.cas.cz/ocas.
129:
130:
131: \section{RESULTS}
132:
133: Due to the tearing mode instability the current sheet tears into O-type islands
134: (plasmoids) which later on coalesce into larger ones. During these processes
135: both the electrons and positrons are accelerated. Figure 1 shows an evolution
136: of the electron distribution function $f(E)$ (for Case I) in dependence on the
137: energy $E$ in the whole numerical plane (upper part of Figure 1) as well as in
138: one selected location (bottom part of Figure 1, for the location see Figure 4),
139: where much harder spectrum can be seen. While the spectral index at
140: $\omega_{pe} t$ = 600 in the whole plane is -3.3, in the selected location is
141: about -1.9. There are other such places, especially at the locations where
142: high-energy electrons and positrons are produced, see Figure 4. This
143: acceleration process is very efficient and looks to have a 2-step character in
144: time as shown in Figure 2, where an evolution of the number of accelerated
145: electrons in the ratio to the total number of electrons (in percents) with the
146: energy $E/m^2 = \gamma
147: >$ 4 is shown. An analysis shows that the first step of the acceleration (up to
148: about $\omega_{pe} t$ = 400) is connected with tearing processes and in the
149: second step (after $\omega_{pe} t$ = 400) the main process is the coalescence
150: of plasmoids. The acceleration with the higher value of
151: $\omega_{ce}/\omega_{pe}$ = 4 (full line) (Case I) is more efficient than that
152: with $\omega_{ce}/\omega_{pe}$ = 1.9 (dashed line) (Case II) (Figure 2).
153:
154: Similarly as in previous studies (Drake et al. 2005, Pritchett 2006, Jaroschek
155: et al. 2004a,b) the electrons and positrons are accelerated in the electric
156: field near the X-points formed during the tearing and coalescence processes.
157: But we found that during this acceleration process (Figures 3 and 4) the
158: electrons and positrons are moving into different locations around the O-type
159: magnetic structures (plasmoids) and thus they are spatially separated (e.g. see
160: the region around $x$ = 500$\Delta$ and $y$ = 450$\Delta$ in Figure 4). To
161: understand this separation process we analyzed the electric field near the
162: X-point of the reconnection (Figures 5 and 6). As seen here the electrons (the
163: asterisks) are located at the borders of the areas with the enhanced ($-E_y$
164: and $-E_x$) and ($+E_y$ and $+E_x$). On the other hand, the positrons are
165: located along the remaining two borders (see Figure 3, bottom part). The
166: electric component $E_z$ along the line $y$ = 150$\Delta$ is negative between
167: the O-type magnetic structures (Figure 6). But only near the X-point of the
168: reconnection this electric field ($E_z$) deviates from that of the inductive
169: one $-{\bf v \times B}/c$ (where ${\bf v}$ is the plasma velocity), see Figure
170: 7, where their profiles are shown at two times ($\omega_{pe} t$ = 50 and 100).
171: This deviation defines the diffusion region of the reconnection. As concerns
172: the magnetic field in the early stage of the reconnection, the magnetic field
173: component $B_y$ is positive in the region $x
174: >$ 1000$\Delta$ and negative for $x <$ 1000$\Delta$. The structure of the magnetic
175: field together with the plasma velocity pattern in the regions from the X-point
176: to the magnetic island centers resembles to that of the collapsing magnetic
177: trap (Giuliani et al. 2005; Karlick\'y \& B\'arta 2006). In such a structure
178: the particles are also accelerated, but not separated as found here near the
179: X-points. This additional acceleration process is known as that in the
180: contracting magnetic islands (Drake et al. 2006).
181:
182:
183: Furthermore, we made the similar computations, but without the guiding magnetic
184: field component ($B_z= 0$). In this case no separation of electrons and
185: positrons was found. Also the electric field structure was different from that
186: presented in Figures 5 and 6.
187:
188: Finally, we made similar computations also for the electron-"proton" plasma,
189: with the electron-proton ratio $m_i/m_e$ = 16 (Case III). Similarly as in the
190: previous cases, Figure 8 shows that accelerated electrons and protons move to
191: different positions. Comparing this Case III (dotted line in Figure 2) with
192: Case I (full line) the number of accelerated electrons is reduced.
193:
194:
195:
196:
197: \section{DISCUSSION AND CONCLUSIONS}
198:
199: The present simulations show that the magnetic reconnection with the guiding
200: magnetic field accelerate electrons to different positions around the plasmoid
201: than for positrons or "protons". If the magnetic field connectivity (in the
202: $z$-direction) from the upper and bottom part of the plasmoid differs then the
203: accelerated electrons and positrons (or protons) move into quite different
204: locations as observed by RHESSI. The separation of particles with different
205: electric charge is a natural consequence of the acceleration in direct electric
206: field near the X-points of the magnetic field reconnection.
207:
208: In agreement with Litvinenko (1996) and Zharkova \& Gordovskyy (2004) we found
209: that the separation process is due to a presence of the non-zero guiding
210: magnetic field (non-neutral current sheet). Namely, our simulations with the
211: zero guiding magnetic field show no such separations.
212:
213: Zharkova \& Gordovskyy (2004) have shown that the separation direction (in the
214: present designation of the electric and magnetic field components) depends on
215: the sign of the term $q^3 B_y B_z E_z$, where $q$ is the electron (-e) or
216: positron (+e) charge (see the relation (8) in their paper). Considering the
217: direction of the magnetic and electric fields in our case (Figures 5 and 6) it
218: can be shown the separation direction found agrees to this relation. This
219: result agrees also to the relations presented in the paper by Litvinenko
220: (1996).
221:
222: The similar separation process is found also for the reconnection in the
223: "proton"-electron plasma with the proton-electron mass ratio $m_i/m_e$ = 16.
224: This mass ratio is not realistic and is taken due to computer limitations.
225: Nevertheless, we expect that such a separation process will be confirmed by
226: future computations also for the real proton-electron mass ratio.
227:
228: Comparing the acceleration process for the electron-positron (full line) and
229: electron-proton ($m_i/m_e$=16, dotted line) plasma in Figure 2, we found that
230: N$_{m_i/m_e=1}$/N$_{m_i/m_e=16}$ is 1.18, i.e. the number of accelerated
231: electrons N depends on the proton-electron mass ratio as N$_{1}$/N$_{m_i/m_e}$
232: $\simeq$ 1/($m_i/m_e$)$^{0.0625}$. If this relation is valid also for the real
233: proton-electron mass ratio then we can write N$_{1}$/N$_{1838}$ = 1.6, which
234: gives enough accelerated electrons also for the real electron-proton plasma.
235:
236: Comparing the present modelling with previous studies the most similar
237: simulation is that of Zenitani \& Hoshino (2001), especially due to initial
238: high thermal plasma energy. But in their model no guiding magnetic field, which
239: is crucial for the particle separation, is considered. Furthermore, contrary to
240: our start from noise level they initiate the reconnection by magnetic field
241: perturbation which can influence the separation process, too. The maximum
242: energies of accelerated electrons in both models are comparable. But in our
243: model the reconnection process is about three times slower than that in
244: Zenitani \& Hoshino (2001). We think that it is due to the initial magnetic
245: field perturbation.
246:
247: Although the acceleration in the contracting magnetic islands does not separate
248: particles of opposite electric charges this process is important for global
249: acceleration. But this process is even more complicated than presented in the
250: paper by Drake et al. (2006). Namely, not only parallel energy of particles
251: increases due to reflection from the ends of contracting magnetic islands (as
252: described by the relation (1) in Drake et al. (2006)) but also the
253: perpendicular energy of particles $E_{\perp}$ can increase due to the betatron
254: type of the acceleration, which follows from the conservation of the magnetic
255: moment $\mu = E_{\perp}/B$ in the region with the increasing magnetic field $B$
256: (see Karlick\'y \& B\'arta 2006). In our simulations the contracting
257: acceleration is time varying, therefore let us compare its efficiency with that
258: of the acceleration near the X-point at one specific time. Using the relation
259: (1) of Drake et al. (2006) we derived the electric field equivalent to this
260: process as $E_{eq}$ = (v$_x$ B$_x^2$ v m$_e$)/($\delta_x$ e B$^2$), where
261: 2$\delta_x$ is the length of the island, $B_x$ and $B$ are the reconnecting and
262: total magnetic fields, $v_x$ is the contracting velocity, $v$ is the mean
263: electron velocity, and $m_e$ is the electron mass. For the parameters in Case I
264: at $\omega_{pe} t$ = 100 it gives the equivalent electric field in the
265: contracting magnetic island $E_{eq}$ one order of magnitude lower than the
266: electric field $E_z$ at the X-point region. It means that at this moment the
267: acceleration near the X-point dominates over that in the contracting magnetic
268: islands.
269:
270:
271: \acknowledgments This research was supported by the Centre for Theoretical
272: Astrophysics and by Grants IAA300030701 of the Academy of Sciences of the Czech
273: Republic. M.~K.\ thanks Prof.~J.\ I.\ Sakai and Dr.~S.\ Saito for many useful
274: discussions concerning numerical modelling. Author also thanks to the referee
275: for constructive comments improving the paper.
276:
277:
278:
279: \begin{thebibliography}{}
280:
281: \bibitem[Bessho \& Bhattacharje (2005)]{BesshoBhattacharje05}
282: Bessho, N., \& Bhattacharje, A. 2005, Phys. Rev. Letters, 95, 245001
283:
284: \bibitem[Coronity (1990)]{Coronity90}
285: Coroniti, F.V. 1990, \apj, 349, 538
286:
287: \bibitem[Drake et al. (2005)]{Drakeetal2005}
288: Drake, J.F., Shay, M.A., Thongthai, W., \& Swisdak, M. 2005, Phys. Rev.
289: Letters, 94, 095001
290:
291: \bibitem[Drake et al. (2006)]{Drakeetal2006}
292: Drake, J.F., Swisdak, M., Che, H.,\& Shay, M.A. 2006, Nature, 443/5, 553
293:
294: \bibitem[Drenkhahn (2002a)]{Drenkhahn02a}
295: Drenkhahn, G. 2002a, \apj, 387, 714
296:
297: \bibitem[Drenkhahn (2002b)]{Drenkhahn02b}
298: Drenkhahn, G. 2002b, \apj, 391, 1141
299:
300: \bibitem[Giuliani et al. (2005)]{Giulianietal05}
301: Giuliani, P., Neukirch, T., \& Wood, P. 2005, \apj, 635, 636
302:
303: \bibitem[Jaroschek et al. (2004a)]{Jaroscheketal04a}
304: Jaroschek, C.H., Lesch, H.,\& Treumann R.A. 2004a, \apj, 605, L9
305:
306: \bibitem[Jaroschek et al. (2004b)]{Jaroscheketal04b}
307: Jaroschek, C.H., Treumann R.A., Lesch, H., \& Scholer, M. 2004b, Physics of
308: Plasmas, 11 (3), 1151
309:
310: \bibitem[Karlick\'y \& B\'arta (2006)]{KarlickyBarta06}
311: Karlick\'y, M., \& B\'arta, M. 2006, \apj, 647, 1472
312:
313: \bibitem[Karlick\'y \& B\'arta (2007)]{KarlickyBarta07}
314: Karlick\'y, M., \& B\'arta, M. 2007, \aap, 464, 735
315:
316: \bibitem[Larrabee et al. (2003)]{Larrabee03}
317: Larrabee, D., Lovelace, R., \& Romanova, M. 2003, \apj, 586, 72
318:
319: \bibitem[Lesch \& Birk (1998)]{LeschBirk98}
320: Lesch, H., \& Birk, G.T. 1998, \apj, 499, 167
321:
322: \bibitem[Litvinenko (1996)]{Litvinenko96}
323: Litvinenko, Yu., E. 1996, \apj, 462, 997
324:
325: \bibitem[Lyubarsky \& Kirk (2001)]{LyubarskyKirk01}
326: Lyubarsky, Y., \& Kirk, J. 2001, \apj, 547, 437
327:
328: \bibitem[Martens \& Young (1990)]{MartensYoung90}
329: Martens, P.C.H., \& Young, A. 1990, \apjs, 73, 333
330:
331: \bibitem[Michel (1994)]{Michel994}
332: Michel, F. 1994, \apj, 431, 397
333:
334: \bibitem[Priest \& Forbes (2000)]{PriestForbes2000}
335: Priest, E.R.,\& Forbes, T. 2000, Magnetic Reconnection: MHD Theory and
336: Applications, Cambridge Univ. Press, Cambridge, UK
337:
338: \bibitem[Pritchett (2006)]{Pritchett2006}
339: Pritchett, P.L. 2006, J. Geophys. Res., 111, A10212
340:
341: \bibitem[Saito \& Sakai (2004)]{SaitoSakai2004}
342: Saito, S., \& Sakai, J.I. 2004, \apj, 616, L179
343:
344: \bibitem[Share et al. (3003)]{Shareetal2003} Share, G.H., Murphy, R.J., Smith, D.M., Lin, R.P., Dennis, B.R.,
345: \& Schwartz, R.A. 2003, \apj, 595, L89
346:
347: \bibitem[Wardle et al. (1998)]{Wardleetal98}
348: Wardle, J., Homan, C., Ojiha, R., \& Roberts, D. 1998, Nature, 395, 457
349:
350: \bibitem[Zenitani \& Hoshino (2001)]{ZenitaniHoshino01}
351: Zenitani, S., \& Hoshino, M. 2001, \apj, 562, L63
352:
353: \bibitem[Zenitani \& Hoshino (2005)]{ZenitaniHoshino05}
354: Zenitani, S., \& Hoshino, M. 2005, \apj, 618, L111
355:
356: \bibitem[Zharkova \& Gordovskyy]{Zharkova2004}
357: Zharkova, V.V.,\& Gordovskyy, M. 2004, \apj, 604, 884
358:
359: \bibitem[Zhu \& Parks (1993)]{ZhuParks93}
360: Zhu, Zh., \& Parks, G. 1993, J. Geophys. Res., 98, 7603
361:
362:
363:
364:
365: \end{thebibliography}
366:
367: \clearpage
368:
369:
370: %% This example uses \plotone to include an EPS file scaled to
371: %% 80% of its natural size with \epsscale. Its caption
372: %% has been written to indicate that additional figure parts will be
373: %% available in the electronic journal.
374:
375:
376:
377: \begin{figure}
378: \epsscale{0.5}
379: \plotone{f1a.ps} \plotone{f1b.ps}
380: \caption{The energy distribution of electrons at the initial state (dotted line),
381: at $\omega_{pe} t$ = 600 (dashed line), and at $\omega_{pe} t$ = 1000 (full line) (Case I).
382: The upper part: The distributions in the whole numerical plane. The bottom part:
383: The distributions in the selected location, in the circle centered
384: at $x$=550$\Delta$ and $y$=450$\Delta$ with the radius $r$ = 50$\Delta$.
385: For this location, see Figure 4.}
386: \end{figure}
387:
388:
389:
390: \begin{figure}
391: \epsscale{1.}
392: \plotone{f2.ps}
393: \caption{The number of accelerated electrons with $E/mc^2$ $>$ 4 expressed
394: in the ratio to total electron number for Case I (full line), Case II (dashed line), and
395: Case III (dotted line).}
396: \end{figure}
397:
398:
399:
400: \begin{figure}
401: \epsscale{0.5}
402: \plotone{f3a.ps} \plotone{f3b.ps}
403: \caption{The spatial distribution of accelerated electrons (dots in upper part)
404: and positrons (dots in bottom
405: part) with
406: the energy $E/mc^2$ $>$ 10
407: superposed on the magnetic field lines projected to the $x-y$ plane
408: at $\omega_{pe} t$ = 150 (Case I). Compare locations of electrons and positrons.}
409: \end{figure}
410:
411:
412:
413: \begin{figure}
414: \epsscale{0.5}
415: \plotone{f4a.ps} \plotone{f4b.ps}
416: \caption{The spatial distribution of accelerated electrons (dots in the upper part)
417: and positrons (dots in the bottom part) with the energy $E/mc^2$ $>$
418: 13 superposed on the magnetic field lines projected
419: to the $x-y$ plane at $\omega_{pe} t$ = 600 (Case I). For the separation of electrons and
420: positrons, see the region around $x$ = 500$\Delta$ and $y$ = 450$\Delta$.}
421: \end{figure}
422:
423:
424:
425:
426: \begin{figure}
427: \epsscale{1.}
428: \plotone{f5.ps}
429: \caption{The spatial distributions of the electric field components ({\bf E}/B$_0$)
430: around the
431: X-point in the current sheet near $y$ = 150$\Delta$ at $\omega_{pe} t$ = 100
432: (Case I), compare with Figure 6.
433: The asterisks mean accelerated electrons for the energy $E/mc^2 >$ 10. The thick full
434: contour means the area of enhanced $\mid E_z \mid$ component (level $E_z/B_0$=-0.22).
435: (Remark: In the upper current sheet with oppositely oriented current along the line
436: $y$=450$\Delta$ the $E_z$ component is oppositely
437: oriented, i.e. positive.)
438: The thick dashed
439: contour is the area of the enhanced $E_y$ (level $E_y/B_0$=0.36). The thick dotted contour
440: means the area with oppositely oriented $E_y$ (level $E_y/B_0$=-0.36). The thin dashed
441: contour is the area of the enhanced $E_x$ (level $E_x/B_0$=0.07). The thin dotted contour
442: means the area with oppositely oriented $E_x$ (level $E_x/B_0$=-0.07).}
443: \end{figure}
444:
445:
446:
447: \begin{figure}
448: \epsscale{1.}
449: \plotone{f6.ps}
450: \caption{The electric field components ($E_x$ (dashed line), $E_y$ (dotted line),
451: and $E_z$ (full line)) along the center of the current sheet, i.e. along the line
452: $y$=150$\Delta$ at $\omega_{pe} t$ = 100 (Case I) (compare with Figure 5).}
453: \end{figure}
454:
455:
456: \begin{figure}
457: \epsscale{1.}
458: \plotone{f7.ps}
459: \caption{The electric field component $E_z$ (full line) and the term
460: $-{\bf v \times B}/c$ (dashed line) at two times (Case I): at $\omega_{pe} t$ =
461: 50 (upper part), and at $\omega_{pe} t$ = 100 (bottom part).}
462: \end{figure}
463:
464:
465:
466: \begin{figure}
467: \epsscale{0.5}
468: \plotone{f8a.ps} \plotone{f8b.ps}
469: \caption{The spatial distribution of accelerated electrons (dots in the upper part) with
470: the energy $E/m_ec^2$ $>$ 5
471: and "protons" (dots in the bottom part) with
472: $E/m_ic^2$ $>$ 1.4
473: near the X-point of the magnetic field structure in the $x-y$ plane
474: at $\omega_{pe} t$ = 50 (Case III). The proton-electron mass ratio is 16.}
475: \end{figure}
476:
477:
478: \end{document}
479: