astro-ph0608586/ms.tex
1: %\documentclass{emulateapj}
2: \documentclass[12pt,preprint]{aastex}
3: %\documentclass[manuscript]{aastex}
4: %\documentclass[preprint2]{aastex}
5: 
6: \shorttitle{Magnetic field in supernova remnant SN~1987A}
7: \shortauthors{E.G. Berezhko \& L.T. Ksenofontov}
8: 
9: \begin{document}
10: 
11: \title {Magnetic field in supernova remnant SN~1987A}
12: 
13: \author{E.G.~Berezhko\altaffilmark{1}
14:         L.T.~Ksenofontov\altaffilmark{1}
15: }
16: \altaffiltext{1}{Yu.G. Shafer Institute of Cosmophysical Research and Aeronomy,
17:                      31 Lenin Ave., 677980 Yakutsk, Russia}
18: 
19: \email{berezhko@ikfia.ysn.ru}
20: 
21: \begin{abstract} 
22: A nonlinear kinetic theory of cosmic ray (CR) acceleration in
23: supernova remnants is employed to investigate the properties of
24: the remnant 
25: SN~1987A. It is shown that a
26: large downstream magnetic field $B_d\approx 10$~mG is required
27: to fit the existing observational data.
28: Such a strong field together with the strong shock modification due to
29: CR backreaction provides the steep 
30: and concave radioemission
31: spectrum and
32: considerable synchrotron cooling of high energy electrons which diminish
33: their X-ray synchrotron flux below the observed Chandra flux
34: which has to be considered as an upper limit for nonthermal
35: X-ray emission. The expected $\gamma$-ray energy flux at TeV-energies at the current epoch 
36: is $2\times 10^{-13}$~erg/(cm$^2$s).
37: \end{abstract}
38: 
39: \keywords{(ISM:) cosmic rays -- acceleration of particles -- shock waves --
40: (stars:) supernovae: individual (SN~1987A) -- radiation mechanisms: non-thermal --
41: X-rays: individual (SN~1987A)} 
42: 
43: %________________________________________________________________
44: 
45: \section{Introduction}
46: 
47: Supernova (SN) 1987A occurred in the Large Magellanic Cloud.
48: It has been extensively studied in all wavelengths from radio to $\gamma$-ray.
49: The initial short outburst of radio emission \citep{turtle87} is attributed
50: to the synchrotron emission of electrons accelerated by the SN shock propagated
51: in the free wind of
52: presupernova star, which was the blue supergiant (BSG) 
53: \citep{chf87}. After about 3 years
54: radio emission was detected  again \citep{ss92, gae97}
55: as well as the monotonically increased X-ray emission \citep{gro94, has96}.
56: This second increase of emission is attributed to
57: the entrance of the outer SN shock into the thermalized BSG wind and then
58: in the H~II
59: region occupied by much more dense matter, consists of the 
60: swept up wind of red
61: supergiant (RSG) progenitor star \citep{chd95}.
62: 
63: To describe the observed properties of nonthermal emission of
64: SN~1987A we use here a nonlinear
65: kinetic theory. It couples the particle acceleration process
66: with the hydrodynamics of the thermal gas \citep{byk96, bv00}.
67: Therefore
68: in spherically symmetric approach it is able to quite definitely
69: predict the evolution of gas density, pressure, mass velocity,
70: together with the energy spectrum
71: and the spatial distribution of cosmic ray (CR) 
72: nuclei and electrons at any given
73: evolutionary epoch $t$, and the properties of the nonthermal radiation
74: produced in SNRs due to these accelerated CRs. The application of this theory
75: to individual SNRs \citep[see][for a review]{ber05} 
76: has provided the 
77: explanation of the observed SNR properties 
78: under the assumption of strong interior 
79: magnetic field. It was also shown that this strong field leads to the concentration of
80: the highest-energy electrons in a very thin shell just behind the shock.
81: Recent observations with the Chandra and XMM-Newton X-ray telescopes in space
82: make it possible to resolve such filamentary structures, which are the 
83: result of strong synchrotron losses of the emitting multi-TeV electrons in
84: strong magnetic fields downstream of the outer SN shock
85: \citep{ber05}.  
86: 
87: Application of the nonlinear theory to SN~1987A \citep{bk00}
88: explained the properties of the observed radioemission. It predicted also the
89: continuous increase of radioemission with roughly constant rate during the period
90: from 1999 to 2006 yr, which was confirmed by further measurements \citep{manch02}.
91:  
92: We analyze here radio data for more extended period of observation
93: \citep{mg01,manch02} together with the Chandra
94: measurements of X-ray emission \citep{park04} which 
95: provides a strong constraint on the amount of
96: the nonthermal synchrotron emission at the same energies 0.5-10~keV. It is
97: demonstrated that this constrain gives the evidence of strongly enhanced
98: (amplified) magnetic field inside the SNR, that in turn provides a better fit of
99: radio data than in our previous study.
100: 
101: \section{Physical parameters of SNR 1987A}
102: We use canonical values of stellar ejecta mass
103: $M_{ej}=10M_{\odot}$, distance $d=50$~kpc, 
104: hydrodynamic explosion energy $E_{sn}=1.5\times 10^{51}$~erg
105: \citep[e.g.][]{mccray93}.
106: During an initial period the shell material has a broad distribution in
107: velocity $v$. The fastest part of this ejecta distribution can be described by
108: a power law $dM_{ej}/dv\propto v^{2-k}$. We use
109: a value $k=8.6$ appropriate for SN~1987A \citep{mccray93}.
110: The interaction of the ejecta with the circumstellar medium 
111: (CSM) creates
112: a strong shock there which heats the thermal gas and accelerates particles
113: diffusively to a nonthermal CR component of comparable energy density. 
114: 
115: Strongly nonuniform CSM is a result of interaction of
116: progenitors winds: dense slow RSG wind  and subsequent fast and
117: diluted BSG wind. 
118: During initial period after the explosion 
119: the SN shock propagated in the free BSG wind and reached after about 
120: a day $\sim 1000$
121: the termination shock in BSG wind, situated at the radial distance
122: $R_T=3.1\times 10^{17}$~cm \citep{bk00}. 
123: After that the SN shock propagates in the
124: thermalized BSG wind of density $\rho_B$. Considerable rapid decrease of the SN
125: shock speed occurred during the days $1500-2500$ \citep{gae97} 
126: showed that SN shock entered the H~II region
127: of density $\rho_R\gg \rho_B$. Therefore we model
128: the CSM density distribution at distances $r>R_T$ in the form
129: %
130: \begin{equation}
131: \rho_0=\frac{\rho_{B}+\rho_{R}}{2}-
132: \frac{\rho_{B}-\rho_{R}}{2}
133: \tanh\frac{r-R_C}{l_{C}},
134: \end{equation}
135: %
136: where $R_C=6.1\times 10^{17}$~cm is the contact discontinuity between two winds,
137: $l_{C}=0.05R_C$ is the scale of the smooth transition between them.
138: 
139: The analysis of the rapidly increasing X-ray emission provides the evidence that
140: H~II region in turn is not uniform: at day $\sim 6200$ the SN shock begins to
141: interact with the dense inner ring. Its radial behavior can be represented
142: in the form 
143: %
144: \begin{equation}
145: \rho_R=\frac{\rho_{R1}+\rho_{R2}}{2}-
146: \frac{\rho_{R1}-\rho_{R2}}{2}
147: \tanh\frac{r-R_R}{l_R},
148: \end{equation}
149: %
150: where $R_R=6.8\times 10^{17}$~cm, $l_R=0.08R_R$, $\rho_{R2}\approx 20
151: \rho_{R1}$. Such a gas radial profile is close to what was extracted
152: from the above analysis \citep{park04}.
153: We adopt here for the gas number density
154: $N_g=\rho/m_p$ the following values:
155: $N_g^B=0.29$~cm$^{-3}$, $N_g^{R1}=280$~cm$^{-3}$
156: and $N_g^{R2}=4000$~cm$^{-3}$. Here $m_p$ is the mass of proton.
157: The value $N_g^{R1}=280$~cm$^{-3}$
158: which is by a factor
159: of 1.5 lower than was used in our previous study \citep{bk00},
160: provides a good compromise between the SN shock dynamics seen in radio and
161: in X-ray emissions (see below). Note, that the actual structure of the H~II
162: region, especially the dense inner ring, is essentially nonspherically
163: symmetric. Nevertheless since our spherically symmetric theory
164: reproduces the SN shock evolution $R_s(t)$ consistent with the existing data,
165: one can expect that it gives a good estimate for the production of CRs and their
166: subsequent nonthermal emission.
167: 
168: A rather high downstream magnetic field strength
169: $B_d\sim 1$~mG is
170: needed to reproduce the observed steep radio spectrum
171: \citep{bk00}. We believe
172: that the required strength of the magnetic field
173: have to be attributed to
174: nonlinear field amplification at the SN shock by CR
175: acceleration itself. According to plasma physical
176: considerations \citep{lb00,bell04}, the existing CSM magnetic field can
177: indeed be significantly amplified at a strong shock by CR
178: streaming instabilities.
179: In fact, for all the
180: thoroughly studied young SNRs, the ratio of magnetic field
181: energy density $B_0^2/8\pi$ in the upstream region of the shock precursor to
182: the CR pressure $P_c$ is about the same
183: \citep{vbk05}. Here $B_0=B_\mathrm{d}/\sigma$ is the far upstream field
184: presumably amplified by CRs of highest energy, $\sigma$ is the total shock compression ratio.
185: Within an error of about 50 percent we have
186: %
187: $ %\begin{equation}
188: B_0^2/(8\pi P_c) \approx 5\times 10^{-3}.
189: $ %\end{equation}
190: %
191: CR pressure in young SNRs has a typical value $P_c\approx 0.5 \rho_0 V_s^2$,
192: therefore we adopt here upstream magnetic field 
193: %
194: \begin{equation}
195: B_0=\sqrt{2\pi \times 10^{-3}\rho_0 V_s^2}.
196: \end{equation}
197: %
198: Since the process of magnetic field amplification is 
199: not included in our theory
200: we simply postulate the existence of far upstream field $B_0$ given by Eq.(3).
201: The spectrum of CRs produced by strong
202: modified shock is very hard so that CRs with highest energies have a largest
203: contribution in their energy content. These the most energetic CRs produce
204: field amplification on their spacial scale that is the precursor size.
205: Therefore CRs with lower energies already 'see' the amplified field $B_0$. 
206: We will argue that such a high field,
207: that is by a factor of ten larger than used in our previous study \citep{bk00},
208: is indeed required to fit the X-ray data. It also gives a better fit of radio
209: data.
210: 
211: We start our consideration from the SNR evolutionary epoch $t=1000$~d, when the
212: outer SN shock has a radius $R_i=R_T$ and speed $V_i=28000$~km/s. These values
213: of $R_s$ and $V_s$ according
214: to our calculations \citep{bk00} correspond to the end of SN
215: shock propagation in the free BSG wind region $r<R_T$. We neglect the
216: contribution of CRs accelerated in the region $r<R_T$, because due to a high gas
217: density the number of CRs produced in the region $r>R_T$ very soon becomes
218: dominant.
219: 
220: \section{Results and Discussion}
221: 
222: Calculated shock radius $R_s$ and speed 
223: $V_s$ shown in Fig.\ref{f1}a as a function of
224: time are in satisfactory agreement with the values obtained on the basis of
225: radio and X-ray measurements. Note that radio data 
226: compared with X-ray data gives larger shock size
227: at any given time $t$ and calculated radius 
228: $R_s(t)$ goes between these two sets
229: 
230: To fit the spectral shape of the observed
231: radio emission we assume a proton injection rate
232: $\eta=3\times 10^{-3}$, which is a fraction of gas particles involved into the
233: acceleration at the from of SN shock. This leads to a significant nonlinear modification of
234: the shock: as it is seen in Fig.\ref{f1}b  total shock compression ratio
235: $\sigma\approx 5.3$ is essentially larger
236: and a subshock compression ratio $\sigma_s\approx 2.8$ is lower 
237: than classical value 4.
238: 
239: Since the SN~1987A is very young CRs despite of very efficient acceleration
240: accumulated only 7\% of the explosion energy $E_{sn}$ (see Fig.\ref{f1}c). 
241: Note that the injection is expected to be strongly suppressed at the 
242: quasiperpendicular part of the
243: shock, therefore
244: one should renormalize the results for the nucleonic spectrum, 
245: calculated within
246: the spherically symmetric model. The lack of symmetry in the actual 
247: SNR can be approximately
248: taken into account by a renormalization factor 
249: $f_{re}\approx 0.2$, which diminishes the nucleonic CR production efficiency, 
250: calculated in the spherical
251: model, and all effects associated with it \citep[see][for details]{vbk03}.
252: With  this renormalization the CRs inside SN~1987A
253: already contain
254: %
255: \begin{equation}
256: E_c=0.07f_{re}E_{sn}\approx 2\times 10^{49}~\mbox{erg}.
257: \end{equation}
258: % 
259: 
260: Since during the last 15~yrs the shock speed 
261: and the gas density are almost constant CR
262: energy content grows roughly linearly with time. It gives the natural
263: explanation of the linear increase of the radio emission detected during this
264: evolutionary period.
265: 
266: Strongly modified SN shock generates CR spectrum 
267: $N\propto p^{-\gamma}$, which is very steep at momenta
268: $p<m_pc$, with index $\gamma =(\sigma_s+2)/(\sigma_s-1)\approx 2.7$.
269: CR electrons with such a spectrum produces synchrotron radioemission
270: spectrum $S_{\nu}\propto \nu^{-\alpha}$ with spectral index $\alpha
271: =(\gamma-1)/2\approx 0.9$, that very well corresponds to the experiment,
272: as it is seen in Fig.\ref{f2}, where we present synchrotron energy spectra
273: $\nu S_{\nu}$,
274: calculated for five subsequent epoch together with the experimental data.
275: Note that CR spectrum has a concave shape: it becomes flatter at higher momenta
276: $p$.
277: As a consequence synchrotron spectrum $S_{\nu}(\nu)$ is also concave as it is
278: clearly seen in Fig.\ref{f2} at $\nu<10^{12}$~Hz. Radio data reveal this property in
279: good consistency with theoretical prediction.
280: 
281: Strong downstream magnetic field $B_d\approx 15$~mG, that corresponds to the
282: upstream field $B_0\approx 3$~mG (see Fig.\ref{f1}), provides synchrotron cooling
283: of electrons with momenta $p>10m_pc$ that in turn makes 
284: synchrotron spectrum at high frequencies $\nu>10^{12}$~Hz
285: very steep (see Fig.\ref{f2}). 
286: Concave shape of electrons continuously produced at the shock front together with
287: their synchrotron cooling lead to a formation of two peaks in synchrotron energy
288: spectrum $\nu S_{\nu}$. The first one at $\nu \approx 10^{12}$~Hz corresponds to
289: CR electron momentum $p\approx 10m_pc$ above which synchrotron energy looses
290: are relevant, whereas the second peak at $\nu \approx 10^{18}$~Hz corresponds
291: to the maximum momentum $p\approx 10^4m_pc$ of accelerated electrons.
292: Under this condition calculated
293: synchrotron flux at frequency $\nu\approx 10^{17}$~Hz, which
294: corresponds to the photon energy $\epsilon_{\gamma}=0.5$~keV, 
295: is below the measured
296: flux at the epochs $t>3000$~d.
297: Since the contribution of the nonthermal radiation in the
298: observed X-ray emission of SN~1987A is not 
299: very well known \citep[e.g.][]{mich02},
300: the observed X-ray flux has to be considered as 
301: the upper limit for the expected nonthermal emission. 
302: At early epoch $t<2500$~d however the calculated flux
303: exceeds the measured one (see the curve, corresponding $t=1970$~d in
304: Fig.\ref{f2}).
305: This can be considered as indication, that the actual magnetic field
306: $B_0$ is few times larger then given by the Eq.(3).
307: 
308: To illustrate the situation expected at considerably lower magnetic field we
309: present in Fig.\ref{f3} synchrotron energy spectra calculated at the same set of
310: parameters as before except magnetic field, which was taken $B_0=200$~$\mu$G
311: independent of time. Since the interior magnetic field $B_d \approx 2$~mG 
312: ($\sigma \approx 10$) is
313: essentially lower in this case, synchrotron losses of high energy CR electrons
314: are considerably smaller compared with the previous case. Due to this fact
315: synchrotron spectra considerably exceeds at any given epoch the measured Chandra
316: flux. Therefore we can conclude, that the actual interior magnetic field
317: strength is not lower than $5$~mG.
318: 
319: Projected radial profile of the nonthermal X-ray emission at high
320: interior magnetic field $B_d=10$~mG is characterized by an extremely 
321: sharp peak of thickness
322: $L\approx 10^{-3}R_s$ just behind the SN shock. Since the thermal X-rays have much
323: wider radial distribution it provides the possibility 
324: for experimental determination
325: of the nonthermal X-ray contribution in the observed emission.
326: Experimental determination of the X-ray spectrum at $\nu =10^{18}-10^{19}$~Hz
327: ($\epsilon_{\gamma}=5-50$~keV) is the another possibility to discriminate the
328: nonthermal emission. Such kind of measurements can be done with space
329: instrument Suzaku.
330: 
331: Completely different 
332: possibility to have the synchrotron flux, which fit the radio data and
333: goes below Chandra data, is low field scenario, when the
334: magnetic field is
335: as low as $B_0<2$~$\mu$G. In such a case the  
336: cutoff frequency of synchrotron
337: spectrum $\nu_{max}$ is lower than $10^{17}$~Hz and 
338: calculated fluxes go down exponentially at $\nu>\nu_{max}$
339: below Chandra points. However the value $B_0<2$~$\mu$G is unrealistically small
340: for such a dense medium, which we have in SN~1987A. 
341: In addition the effective energy of electrons
342: %
343: \begin{equation} 
344: \epsilon_e\approx 
345: 5\sqrt{(\nu/\mbox{1~GHz})/(B_d/10~\mu\mbox{G})}\mbox{~~GeV}, 
346: \end{equation}
347: % 
348: which radiate at $\nu \sim 1$~GHz in the interior field $B_d=10$~$\mu$G
349: is about 5~GeV. The electron spectrum at GeV-energies
350: is characterized by a power low index $\gamma\approx 2$ that is considerably
351: smaller than what is required for the observed radio spectrum.
352: Note that as it is seen in Fig.\ref{f3} even the field $B_d\approx 2$~mG is too small
353: to have a good fit of radio data as in Fig.\ref{f2}.
354: Therefore low field scenario should be rejected.
355: 
356: Calculated $\gamma$-ray integral flux shown in Fig.\ref{f4} 
357: at all energies is dominated by the $\pi^0$-decay component.
358: Since the SN shock is strongly modified $\gamma$-ray spectrum at
359: energies $\epsilon_{\gamma}>0.1$~TeV is very hard: $F_{\gamma}\propto
360: \epsilon_{\gamma}^{-0.9}$. 
361: At the current epoch the expected $\gamma$-ray energy flux at
362: TeV-energies is about 
363: $\epsilon_{\gamma}F_{\gamma}\approx 2\times 10^{-13}$~erg/(cm$^2$s) 
364: and during the next four years it
365: expect to grow by a factor of two. This flux is by a factor of five
366: lower compared with our previous prediction \citep{bk00}, because we use here
367: renormalization factor $f_{re} =0.2$, which reduces the number of accelerated
368: CRs calculated within spherically symmetric approach. 
369: The existence of strongly asymmetric CSM structure, which is dense inner ring,
370: makes our prediction of $\gamma$-ray flux uncertain. According to the rough estimate this
371: uncertainty is not very large, about a factor of two, due to the stronger
372: SN shock deceleration in denser medium.
373: At the moment there are only upper limits of TeV emission obtained by CANGAROO
374: \citep{eno03} and HESS \citep{rowell04} instruments (see Fig.\ref{f4}).
375: 
376: \section{Summary}
377: The kinetic nonlinear model for CR acceleration in SNRs has been in
378: detail applied to SN~1987A, 
379: in order to compare its results with 
380: observational properties.We find that quite a reasonable 
381: consistency with most of the observational
382: data can be achieved. 
383: 
384: The evidence of the
385: efficient CR production leading to a strong shock modification 
386: comes from radio data. Significant shock modification
387: leads to the steep and concave CR spectrum 
388: which very well fits the observed radioemission spectrum, under
389: the condition of extremely high downstream magnetic field strength 
390: $B_d\sim 1$~mG. To be consistent also with the Chandra measurements
391: of the X-ray flux, which has to be
392: considered as an upper limit for the nonthermal X-ray emission, even larger
393: interior field $B_d\approx 10$~mG is needed. Such a high field provides
394: a strong synchrotron losses of CR electrons emitting nonthermal
395: X-rays, that makes the high frequency part of the synchrotron spectrum much
396: more steeper and provides a consistency with the experiment.
397: 
398: The expected $\pi^0$-decay $\gamma$-ray energy flux 
399: at the current epoch is about
400: $\epsilon_{\gamma}F_{\gamma}\approx 2\times 10^{-13}$
401: ~erg/(cm$^2$~s) at energies
402: $\epsilon_{\gamma}=0.1-10$~TeV. Therefore the detection of
403: $\gamma$-ray emission at these energies would imply clear evidence 
404: for a hadronic origin and for a strong magnetic field amplification inside
405: SN~1987A.
406: 
407: \begin{thebibliography}{}
408: 
409: \bibitem[Bell(2004)]{bell04}
410: Bell, A. R. 2004, \mnras, 353, 550
411: 
412: \bibitem[Berezhko et al.(1996)]{byk96}
413: Berezhko, E. G., Elshin, V. K., \& Ksenofontov, L. T. 1996, JETP, 82, 1
414: 
415: \bibitem[Berezhko \& V\"olk(2000)]{bv00}
416: Berezhko, E.G. \& V\"olk, H.J. 2000, A\&A, 357, 183
417: 
418: \bibitem[Berezhko \& Ksenofontov(2000)]{bk00}
419: Berezhko, E. G., \& Ksenofontov, L. T. 2000, Astronomy Letters, 26, 639
420: 
421: \bibitem[Berezhko(2005)]{ber05}
422: Berezhko, E.G. 2005, Adv. Space Res., 35, 1031
423: 
424: \bibitem[Chevalier \& Dwarkadas(1995)]{chd95}
425: Chevalier, R. A. \& Dwarkadas, V. V. 1995, \apjl, 452, L45
426: 
427: \bibitem[Chevalier \& Fransson(1987)]{chf87}
428: Chevalier, R. A., \& Fransson, C. 1987, Nature, 328, 44
429: 
430: \bibitem[Enomoto et al.(2003)]{eno03}
431: Enomoto, R. et al.
432: 2003, \apj, 591, L25
433: 
434: \bibitem[Gaensler et al.(1997)]{gae97}
435: Gaensler, B. M. et al. 
436: 1997, \apj, 479, 845
437: 
438: \bibitem[Gorenstein et al.(1994)]{gro94}
439: Gorenstein, P., Hughes, J. P., Tucker, W. H. 1994, \apjl, 420, L25
440: 	
441: \bibitem[Hasinger et al.(1996)]{has96}
442: Hasinger, G., Ashenbach, B. \& Tr\"umper, J. 1996, \aap, 312, L9
443: 
444: \bibitem[Lucek \& Bell(2000)]{lb00}
445: Lucek, S. G. \& Bell, A. R. 2000, \mnras, 314, 65
446: 
447: \bibitem[McCray(1993)]{mccray93}
448: McCray, R. 1993, \araa, 31 175
449: 
450: \bibitem[Manchester et al.(2002)]{manch02}
451: Manchester, R. N. et al. 2002,
452: PASA, 19, 207
453: 
454: \bibitem[Manchester \& Gaensler(2001)]{mg01}
455: Manchester, R. N. \& Gaensler, B. M. 2001, IAU Circ. 7757
456: 
457: \bibitem[Michael et al.(2002)]{mich02}
458: Michael, E. et al. 2002, \apj, 574, 166
459: 
460: \bibitem[Park et al.(2004)]{park04}
461: Park, S. et al. 2004,
462: \apj, 610, 275
463: 
464: \bibitem[Rowell et al.(2004)]{rowell04}
465: Rowell, G. et al. 2004, AIP Conf. Proc., 745,
466: 299
467: 
468: \bibitem[Staveley-Smith et al.(1992)]{ss92}
469: Staveley-Smith, L., et al. 1992,
470: Nature, 355, 147
471: 
472: \bibitem[Turtle et al.(1987)]{turtle87}
473: Turtle, A. J. et al. 1987, Nature, 327, 38
474: 
475: \bibitem[V\"olk et al.(2003)]{vbk03}
476: V\"olk, H. J., Berezhko, E. G., \& Ksenofontov, L. T. 2003, \aap 409, 563
477: 
478: \bibitem[V\"olk et al.(2005)]{vbk05}
479: V\"olk, H. J., Berezhko, E. G., \& Ksenofontov, L. T. 2005, \aap, 433, 229.
480: 
481: \end{thebibliography}
482: 
483: \clearpage
484: 
485: \begin{figure}
486: \epsscale{0.5}
487: \plotone{f1.eps}
488: \caption{(a) Shock radius $R_\mathrm{s}$, 
489: shock speed $V_\mathrm{s}$, 
490: gas density $N_g$ and upstream magnetic field $B_0$ at the current shock
491: position; (b) total
492: shock ($\sigma$) and subshock ($\sigma_\mathrm{s}$) compression ratios;
493: (c) kinetic energy of ejecta $E_{ej}$ and accelerated CR energy content $E_c$
494: as functions of time. 
495: The {\it dotted
496: vertical line} marks the current epoch. The observed radius of
497: the SN shock, as determined by radio \citep{manch02} 
498: and X-ray measurements \citep{park04}, are shown by circles and stars
499: respectively.
500: The scaling values are $R_i=R_T=3.1\times 10^{17}$~cm and $V_i=28000$~km/s
501: \label{f1}}
502: \end{figure}
503: 
504: \clearpage
505: 
506: \begin{figure}
507: \epsscale{1.0}
508: \plotone{f2.eps}
509: \caption{Synchrotron energy spectrum of SN~1987A, calculated for 
510: the five evolutionary epochs. The 
511: {\it ATCA} radio \citep{manch02, mg01}
512: and Chandra X-ray \citep{park04} data for three epochs are shown as
513: well. Higher measured fluxes correspond to later epoch.
514: \label{f2}}
515: \end{figure}
516: 
517: \clearpage
518: 
519: \begin{figure}
520: \plotone{f3.eps}
521: \caption{Same as in Fig.\ref{f2}, but the calculations are for the 
522: case of constant upstream magnetic field $B_0=200$~$\mu$G.
523: \label{f3}}
524: \end{figure}
525: 
526: \clearpage
527: 
528: \begin{figure}
529: \plotone{f4.eps}
530: \caption{Integral $\gamma$-ray energy flux from SN~1987A, calculated for the same five
531: epoch as in Fig.\ref{f2}. CANGAROO \citep{eno03} and HESS \citep{rowell04}
532: upper limits are shown as well.
533: \label{f4}}
534: \end{figure}
535: 
536: \end{document}
537: