1: % Use only LaTeX2e, calling the article.cls class and 12-point type.
2:
3: \documentclass[12pt]{article}
4: \usepackage{epsf}
5:
6: % Users of the {thebibliography} environment or BibTeX should use the
7: % scicite.sty package, downloadable from *Science* at
8: % www.sciencmag.org/misc/con-info.shtml . This package should properly
9: % format in-text reference calls and reference-list numbers.
10:
11: \usepackage{scicite}
12:
13: % Use times if you have the font installed; otherwise, comment out the
14: % following line.
15:
16: \usepackage{times}
17:
18: % The preamble here sets up a lot of new/revised commands and
19: % environments. It's annoying, but please do *not* try to strip these
20: % out into a separate .sty file (which could lead to the loss of some
21: % information when we convert the file to other formats). Instead, keep
22: % them in the preamble of your main LaTeX source file.
23:
24: % personally defined
25: \def\ga{\lower.5ex\hbox{$\; \buildrel > \over \sim \;$}}
26: \def\la{\lower.5ex\hbox{$\; \buildrel < \over \sim \;$}}
27: \def\etal{{\it et al.}}
28: \def\ie{{\it i.e.,~}}
29: \def\mG{~\mu{\rm G}}
30: \def\nG{~{\rm nG}}
31: \def\kms{~{\rm km~s^{-1}}}
32: \def\Mpc{~h^{-1}{\rm Mpc}}
33:
34: % The following parameters seem to provide a reasonable page setup.
35:
36: \topmargin 0.0cm
37: \oddsidemargin 0.2cm
38: \textwidth 16cm
39: \textheight 21cm
40: \footskip 1.0cm
41:
42: % The next command sets up an environment for the abstract to your paper.
43:
44: \newenvironment{sciabstract}{%
45: \begin{quote} \bf}
46: {\end{quote}}
47:
48: % If your reference list includes text notes as well as references,
49: % include the following line; otherwise, comment it out.
50:
51: \renewcommand\refname{References and Notes}
52:
53: % The following lines set up an environment for the last note in the
54: % reference list, which commonly includes acknowledgments of funding,
55: % help, etc. It's intended for users of BibTeX or the {thebibliography}
56: % environment. Users who are hand-coding their references at the end
57: % using a list environment such as {enumerate} can simply add another
58: % item at the end, and it will be numbered automatically.
59:
60: \newcounter{lastnote}
61: \newenvironment{scilastnote}{%
62: \setcounter{lastnote}{\value{enumiv}}%
63: \addtocounter{lastnote}{+1}%
64: \begin{list}%
65: {\arabic{lastnote}.}
66: {\setlength{\leftmargin}{.22in}}
67: {\setlength{\labelsep}{.5em}}}
68: {\end{list}}
69:
70: % Include your paper's title here
71:
72: \title{Turbulence and Magnetic Fields
73: in the Large Scale Structure of the Universe}
74:
75: % Place the author information here. Please hand-code the contact
76: % information and notecalls; do *not* use \footnote commands. Let the
77: % author contact information appear immediately below the author names
78: % as shown. We would also prefer that you don't change the type-size
79: % settings shown here.
80:
81: \author{
82: Dongsu Ryu,$^{1\ast}$ Hyesung Kang,$^{2}$ Jungyeon Cho,$^{1}$
83: and Santabrata Das$^{3}$ \\
84: \\
85: \normalsize{$^{1}$Department of Astronomy and Space Science,
86: Chungnam National University,}\\
87: \normalsize{Daejeon 305-764, Korea}\\
88: \normalsize{$^{2}$Department of Earth Sciences,
89: Pusan National University, Pusan 609-735, Korea}\\
90: \normalsize{$^{3}$Astrophysical Research Center for the Structure and
91: Evolution of the Cosmos,}\\
92: \normalsize{Sejong University, Seoul 143-747, Korea}\\
93: \\
94: \normalsize{$^\ast$To whom correspondence should be addressed.
95: E-mail: ryu@canopus.cnu.ac.kr}
96: }
97:
98: % Include the date command, but leave its argument blank.
99:
100: \date{}
101:
102: %%%%%%%%%%%%%%%%% END OF PREAMBLE %%%%%%%%%%%%%%%%
103:
104: \begin{document}
105:
106: % Double-space the manuscript.
107:
108: \baselineskip24pt
109:
110: % Make the title.
111:
112: \maketitle
113:
114: % Place your abstract within the special {sciabstract} environment.
115:
116: \begin{sciabstract}
117: % 143 words
118:
119: The nature and origin of turbulence and magnetic fields in the
120: intergalactic space are important problems that are yet to be
121: understood.
122: We propose a scenario in which turbulent flow motions are
123: induced via the cascade of the vorticity generated at
124: cosmological shocks during the formation of the large scale structure.
125: The turbulence in turn amplifies weak seed magnetic fields of any
126: origin.
127: Supercomputer simulations show that the turbulence is
128: subsonic inside clusters/groups of galaxies, whereas it is
129: transonic or mildly supersonic in filaments.
130: Based on a turbulence dynamo model, we then estimate that the
131: average magnetic field strength would be a few $\mG$ inside
132: clusters/groups, approximately $0.1 \mG$ around clusters/groups,
133: and approximately $10 \nG$ in filaments.
134: Our model presents a physical mechanism that transfers the gravitation
135: energy to the turbulence and magnetic field energies in the large
136: scale structure of the universe.
137:
138: \end{sciabstract}
139:
140: \clearpage
141:
142: % In setting up this template for *Science* papers, we've used both
143: % the \section* command and the \paragraph* command for topical
144: % divisions. Which you use will of course depend on the type of paper
145: % you're writing. Review Articles tend to have displayed headings, for
146: % which \section* is more appropriate; Research Articles, when they have
147: % formal topical divisions at all, tend to signal them with bold text
148: % that runs into the paragraph, for which \paragraph* is the right
149: % choice. Either way, use the asterisk (*) modifier, as shown, to
150: % suppress numbering.
151:
152: % \paragraph{Introduction.}
153: % 501 words
154:
155: There is growing evidence that the intergalactic medium (IGM) is
156: permeated with magnetic fields and is in a state of turbulence,
157: similar to the interstellar medium within galaxies.
158: Magnetic fields in the intracluster medium (ICM) have been measured
159: using a variety of techniques, including observations of diffuse
160: synchrotron emission from radio halos, inverse-Compton scattered
161: cosmic background radiation in extreme ultraviolet and hard x-ray
162: radiation, and Faraday rotation measure (RM).
163: The inferred strength of the magnetic fields is on the order of
164: $1 \mG$ \cite{ct02,gf04,ckb01}.
165: In the IGM outside of clusters, an upper limit of $\sim 0.1\mG$ has
166: been placed on the magnetic field strength of filaments, based on the
167: observed limit of the RMs of background quasars \cite{rkb98,xkhd06}.
168:
169: So far, signatures of turbulence have been observed only in the ICM.
170: The analysis of the gas pressure maps of the Coma cluster revealed that
171: pressure fluctuations are consistent with Kolmogoroff turbulence,
172: and turbulence is likely to be subsonic with
173: $\varepsilon_{\rm turb} \ga 0.1 \varepsilon_{\rm th}$,
174: where $\varepsilon_{\rm turb}$ and $\varepsilon_{\rm th}$ are
175: the turbulence and thermal energy densities, respectively
176: \cite{sfmb04}.
177: These results agree with predictions of numerical simulations
178: of large scale structure (LSS) formation \cite{kcor97,nvk07}.
179: Turbulence in the ICM also has been studied in RM maps of a few
180: clusters \cite{mgfg04,ve05}.
181:
182: It has been suggested that cosmological shocks
183: with Mach numbers up to $\sim 10^4$ and
184: speeds up to a few thousand ${\rm km~s^{-1}}$
185: exist in the IGM \cite{rkhj03,psej06,krco07}.
186: Such shocks result from the supersonic flow motions that are
187: induced by the hierarchical formation of LSS in the universe.
188: They are collisionless shocks, which form in a tenuous plasma via
189: collective electromagnetic interactions between particles and
190: electromagnetic fields \cite{quest88}.
191: The gravitational energy released during the structure formation
192: is transferred by these shocks to the IGM plasma in several
193: different forms: in addition to the gas entropy,
194: cosmic rays are produced via diffusive shock acceleration
195: \cite{bell78,bo78},
196: magnetic fields are generated via the Biermann battery mechanism
197: \cite{biermann50,kcor97} and Weibel instability \cite{weibel59,msk06},
198: and vorticity is generated at curved shocks \cite{binn74,dw00}.
199:
200: In astrophysical plasmas in which charged particles are coupled to
201: magnetic fields, turbulent flow motions and magnetic fields are
202: closely related.
203: We suggest that the turbulence in the IGM is induced by the
204: cascade of the vorticity generated at cosmological shocks.
205: The turbulence then amplifies the intergalactic magnetic fields
206: (IGMFs) through the stretching of field lines, a process known as
207: the turbulence dynamo.
208: This scenario provides a theoretically motivated model
209: for the evolution of the IGMFs in LSS, independent of the origin of
210: seed fields.
211:
212: There are other sources that can also provide turbulence and magnetic
213: fields to the IGM.
214: For instance, galactic winds can drag out the galactic magnetic fields
215: on the order of $1 \mG$ strength into the surrounding IGM \cite{va99}.
216: The magnetic fields in the lobes of the jets from galactic black holes
217: can also contaminate the IGM \cite{kdlc01}.
218: Mergers of smaller objects are expected to produce turbulent motions
219: in the ICM, which in turn amplify the existing magnetic fields
220: \cite{ssh06}.
221: Those processes, although possibly important, are not topics of
222: this study.
223:
224: % \paragraph{Turbulence.}
225: % 503 words
226:
227: We first calculated the vorticity,
228: ${\vec \omega} \equiv {\vec\nabla}\times{\vec v}$ (curl of flow velocity),
229: in the IGM, from a numerical simulation using particle-mesh/Eulerian
230: hydrodynamic code \cite{rokc93} for the formation of LSS in a cold dark
231: matter dominated universe with a cosmological constant [supporting
232: online material (SOM) text S1].
233: As shown in Fig. 1, numerous shocks exist in the LSS that are bounded by
234: accretion shocks \cite{rkhj03}.
235: The distribution of vorticity closely matches that of shocks,
236: suggesting that a substantial portion of the vorticity, if not all,
237: must have been generated at the shocks.
238:
239: There is a clear trend that the vorticity is larger in hotter
240: (Fig. 2) and denser (fig. S1) regions.
241: As shown in the top right panel of Fig. 2, at the present epoch,
242: $\omega_{\rm rms}t_{\rm age} \sim 10$ to $30$ ($\omega_{\rm rms}$,
243: the root mean square of the vorticity; $t_{\rm age}$, the present
244: age of the universe) in clusters/groups (temperature $T > 10^7$ K)
245: and filaments ($10^5 < T < 10^7$ K), whereas it is on the order of
246: unity in sheetlike structures ($10^4 < T < 10^5$ K) and even smaller
247: in voids ($T < 10^4$ K) (see SOM text S2 for the temperature
248: phases of the IGM).
249: It increases a little with time and asymptotes after redshift $z \la 1$.
250: Because the local eddy turnover time, $t_{\rm eddy}$, can be defined
251: with the vorticity as $t_{\rm eddy} = 1/\omega$,
252: $\omega t_{\rm age}(z)$ represents the number of eddy
253: turnovers in the age of the universe at a given $z$.
254: Roughly, if $\omega t_{\rm age}$ is greater than a few, we expect
255: there has been enough time for the vorticity to cascade down to
256: smaller scales and for turbulence to develop in the IGM.
257: So it is likely that turbulence is well developed in clusters/groups
258: and filaments, but the flow is mostly non-turbulent in sheets and voids.
259:
260: In our simulation the vorticity was generated either directly at
261: curved cosmological shocks or by the baroclinity of flows.
262: The baroclinity resulted from the entropy variation induced at shocks.
263: Therefore, the baroclinic vorticity generation also can be attributed
264: to the presence of cosmological shocks.
265: Our estimates of vorticity generation by the two processes
266: (SOM text S3) are shown with open symbols in the top right panel
267: of Fig. 2.
268: They agree reasonably well with the vorticity present in the
269: simulation, although the estimates are intended to be rough.
270: The plot indicates that the contributions from the two processes
271: are comparable.
272:
273: To estimate the energy associated with turbulence,
274: the curl component of flow motions, ${\vec v}_{\rm curl}$, which
275: satisfies the relation
276: ${\vec\nabla}\times{\vec v}_{\rm curl} \equiv {\vec\nabla}\times{\vec v}$,
277: is extracted from the velocity field (SOM text S4).
278: As vorticity cascades to develop into turbulence, the energy
279: $(1/2)\rho v_{\rm curl}^2$ ($\rho$, gas density) is transferred to
280: turbulent motions, so we regard it as the turbulence energy,
281: $\varepsilon_{\rm turb}$.
282: As shown in Fig. 3,
283: $\varepsilon_{\rm turb} < \varepsilon_{\rm th}$ in clusters/groups.
284: In particular, the mass-averaged value is $\left<\varepsilon_{\rm turb}
285: /\varepsilon_{\rm th}\right>_{\rm mass} = 0.1$ to $0.3$ for $T > 10^7$K,
286: which is in good agreement with the observationally inferred value in
287: cluster cores \cite{sfmb04}.
288: The turbulence Mach number
289: $M_{\rm turb} \equiv v_{\rm turb}/c_{\rm s} = \sqrt{1.8} \
290: (\varepsilon_{\rm turb} / \varepsilon_{\rm th})^{1/2}$,
291: where $c_{\rm s}$ is the sound speed.
292: Therefore, overall turbulence is subsonic in clusters/groups,
293: whereas it is transonic or mildly supersonic in filaments.
294:
295: % \paragraph{Magnetic Fields.}
296: % 704 words
297:
298: The general consensus regarding the origin of the IGMFs is that no
299: mechanism can produce strong coherent magnetic fields in the IGM
300: before the formation of LSS and galaxies \cite{kz07}.
301: However, it is reasonable to assume that weak seed fields were
302: created in the early universe (SOM text S5).
303: The seed fields can be amplified by the intergalactic turbulence
304: discussed above.
305: In principle, if we were to perform magnetohydrodynamic (MHD)
306: simulations of structure formation, the amplification of the IGMFs
307: could be followed.
308: In practice, however, the currently available computational resources
309: do not allow a numerical resolution high enough to reproduce the full
310: development of MHD turbulence in LSS \cite{kcor97}.
311:
312: In order to follow the growth of the IGMFs by the dynamo action of
313: turbulence, we turned to a separate simulation in a controlled box.
314: Starting with a very weak regular field, a three-dimensional
315: incompressible simulation of driven MHD turbulence was performed
316: (SOM text S6).
317: In the simulation, the evolution of magnetic fields goes through
318: three stages:
319: (i) the initial exponential growth stage, when the back-reaction of
320: magnetic fields is negligible;
321: (ii) the linear growth stage, when the back-reaction starts to operate;
322: and (iii) the final saturation stage \cite{cv00}.
323: Adopting the simulation result, we model the growth and saturation
324: of magnetic energy as
325: $$
326: \phi(t/t_{\rm eddy}) = {\varepsilon_B \over \varepsilon_{\rm turb}}
327: = \cases{
328: 0.04 \times \exp\left[(t/t_{\rm eddy}-4)/0.36\right]
329: & for\ $t/t_{\rm eddy}<4$ \cr
330: (0.36/41) \times (t/t_{\rm eddy}-4) + 0.04
331: & for\ $4<t/t_{\rm eddy}<45$ \cr
332: 0.4
333: & for\ $t/t_{\rm eddy}>45$ \cr}
334: \eqno{(1)}
335: $$
336: (fig. S2).
337: Assuming that the fraction of turbulence energy governed by Eq. 1,
338: $\phi$, is converted into the magnetic energy, we estimate the
339: strength of the IGMFs as
340: $B=\left[8\pi\varepsilon_{\rm turb} \cdot \phi(\omega t_{\rm age})
341: \right]^{1/2}$.
342: Here the values of $\omega$ and $\varepsilon_{\rm turb}$ are
343: calculated locally from the structure formation simulation.
344:
345: The resulting IGMFs follows the cosmic web of matter distribution
346: as shown in Fig. 4 (and in fig. S3).
347: On average the IGMFs are stronger in hotter (Fig. 2) and denser
348: (fig. S1) regions in our model.
349: The strength of the IGMFs is $B \ga 1 \mG$ inside
350: clusters/groups (the mass-averaged value for $T > 10^7$ K),
351: $\sim 0.1 \mG$ around clusters/groups (the volume-averaged value for
352: $T > 10^7$ K), and $\sim 10 \nG$ in filaments at present (bottom right
353: panel of Fig. 2) (see SOM text S7 for the numerical convergence
354: of the estimation).
355: These values agree with the observed field strengths discussed earlier.
356: They also agree with the previous study \cite{kcor97}, in which the
357: magnetic field strength in clusters was estimated to be a few $\mu$G,
358: based on a kinetic theory.
359: The IGMFs should be much weaker in sheetlike structures and voids.
360: But as noted above, turbulence is not fully developed in such low
361: density regions, so our model is not adequate to
362: predict the field strength there.
363: For each temperature phase, the IGMFs are stronger in the past,
364: because the gas density is higher.
365: However, the IGMFs averaged over the entire computational volume are
366: weaker in the past because the fraction of strong field regions
367: is smaller.
368:
369: While being amplified, magnetic fields become coherent
370: through the inverse cascade \cite{cv00}.
371: The coherence scale of magnetic fields in fully developed turbulence
372: is expected to be several times smaller than the driving scale,
373: that is, the scale of dominant eddies (SOM text S8).
374: In the IGM outside of clusters, the curvature radius of typical
375: cosmological shocks is approximately a couple of Mpc
376: \cite{rkhj03} (fig. S4), which should represent a characteristic
377: scale of dominant eddies.
378: The coherence length of the IGMFs there is then expected to be
379: several 100 kpc.
380: On the other hand, the scale height of the ICM is several 100 kpc.
381: If it corresponds to the scale of the dominant eddies, the coherence
382: length in the ICM is expected to be $\sim 100$ kpc or so.
383:
384: Our model can predict the RMs owing to the IGMFs, which may be tested
385: in future observations with Low Frequency Array and Square Kilometer
386: Array \cite{gbf04}.
387: Also, our model IGMFs can be employed in the study of the propagation
388: of ultra-high-energy cosmic rays, which is crucial to search for
389: astrophysical accelerators of such high energy particles \cite{dkrc08}.
390:
391: \clearpage
392:
393: % captions 275 words
394:
395: \begin{figure}
396: \vspace{0cm}
397: \centerline{\epsfxsize=16cm\epsfbox{f1.eps}}
398: \vspace{0.5cm}
399: {Fig. 1. Two-dimensional images showing gas density $\rho$ in a
400: logarithmic scale (left), locations of shocks with color-coded
401: shock speed $v_{\rm shock}$ (middle), and magnitude of vorticity
402: $\omega t_{\rm age}$ (right), around a cluster complex of
403: $(25 \Mpc)^2$ area at present ($z=0$).
404: Here, $h$ is the Hubble constant in units of $100 \kms {\rm Mpc}^{-1}$.
405: The complex includes a cluster of x-ray emission-weighted temperature
406: $T_x \approx 3.3$ keV.
407: Color codes for each panel are (left) $\rho/\left< \rho \right>$ from
408: $10^{-1}$ (green) to $10^4$ (red); (middle) $v_{\rm shock}$ from 15
409: (green) to 1,800 $\kms$ (red); and (right) $\omega t_{\rm age}$ from
410: 0.5 (green) to 100 (red).}
411: \end{figure}
412:
413: \clearpage
414:
415: \begin{figure}
416: \vspace{0cm}
417: \centerline{\epsfxsize=16cm\epsfbox{f2.eps}}
418: \vspace{0cm}
419: {Fig. 2. (Left) Volume fraction with given temperature and
420: vorticity magnitude (top left) and temperature and magnetic field
421: strength (bottom left) at present.
422: (Right) Time evolution of the root mean square of the vorticity
423: (top right) and the volume-averaged magnetic field strength
424: (bottom right) for four temperature phases of the IGM and for all
425: the gas as a function of redshift $z$.
426: Magenta symbols in the top right panel are our estimates of the
427: vorticity generated directly at curved shocks (open circles) and
428: by the baroclinity of flows (open squares).
429: Magenta open circles in the bottom right panel show the mass-averaged
430: magnetic field strength for $T > 10^7$ K.}
431: \end{figure}
432:
433: \clearpage
434:
435: \begin{figure}
436: \vspace{-4cm}
437: \centerline{\epsfxsize=16cm\epsfbox{f3.eps}}
438: \vspace{-2.5cm}
439: {Fig. 3. Ratio of turbulence to thermal energies as a function of
440: temperature at present.
441: The values shown are volume-averaged and mass-averaged over
442: temperature bins.}
443: \end{figure}
444:
445: \clearpage
446:
447: \begin{figure}
448: \vspace{0cm}
449: \centerline{\epsfxsize=11cm\epsfbox{f4.eps}}
450: \vspace{0.5cm}
451: {Fig. 4. Volume-rendering image showing the logarithmically scaled
452: magnetic field strength at $z=0$ in the whole computational box of
453: $(100 \Mpc)^3$ volume.
454: Color codes the magnetic field strength from $0.1 \nG$ (yellow) to
455: $10 \mG$ (magenta).
456: The colors were chosen so that clusters and groups show as
457: magenta and blue and filaments as green.}
458: \end{figure}
459:
460: \clearpage
461:
462: % references 422 words
463:
464: \begin{thebibliography}{}
465:
466: \bibitem{ct02}
467: C. L. Carilli, G. B. Taylor,
468: {\it Annu. Rev. Astron. Astrophys.} {\bf 40}, 319 (2002)
469:
470: \bibitem{gf04}
471: F. Govoni, L. Feretti,
472: {\it Int. J. Mod. Phys. D} {\bf 13}, 1549 (2004)
473:
474: \bibitem{ckb01}
475: T. E. Clarke, P. P. Kronberg, H. B\"ohringer,
476: {\it Astrophys. J.} {\bf 547}, L111 (2001)
477:
478: \bibitem{rkb98}
479: D. Ryu, H. Kang, P. L. Biermann,
480: {\it Astron. Astrophys.} {\bf 335}, 19 (1998)
481:
482: \bibitem{xkhd06}
483: Y. Xu, P. P. Kronberg, S. Habib, Q. W. Dufton,
484: {\it Astrophys. J.} {\bf 637}, 19 (2006)
485:
486: \bibitem{sfmb04}
487: P. Schuecker, A. Finoguenov, F. Miniati, H. B\"ohringer, U. G. Briel,
488: {\it Astron. Astrophys.} {\bf 426}, 387 (2004)
489:
490: \bibitem{kcor97}
491: R. M. Kulsrud, R. Cen, J. P. Ostriker, D. Ryu,
492: {\it Astrophys. J.} {\bf 480}, 481 (1997)
493:
494: \bibitem{nvk07}
495: D. Nagai, A. Vikhlinin, A. V. Kravtsov,
496: {\it Astrophys. J.} {\bf 655}, 98 (2007)
497:
498: \bibitem{mgfg04}
499: M. Murgia \etal,
500: {\it Astron. Astrophys.} {\bf 424}, 429 (2004)
501:
502: \bibitem{ve05}
503: C. Vogt, T. En{\ss}lin,
504: {\it Astron. Astrophys.} {\bf 434}, 67 (2005)
505:
506: \bibitem{rkhj03}
507: D. Ryu, H. Kang, E. Hallman, T. W. Jones,
508: {\it Astrophys. J.} {\bf 593}, 599 (2003)
509:
510: \bibitem{psej06}
511: C. Pfrommer, V. Springel, T. A. En{\ss}lin, M. Jubelgas,
512: {\it Mon. Not. Roy. Astron. Soc.} {\bf 367}, 113 (2006)
513:
514: \bibitem{krco07}
515: H. Kang, D. Ryu, R. Cen, J. P. Ostriker,
516: {\it Astrophys. J.} {\bf 669}, 729 (2007)
517:
518: \bibitem{quest88}
519: K. B. Quest,
520: {\it J. Geophys. Res.} {\bf 93}, 9649 (1988)
521:
522: \bibitem{bell78}
523: A. R. Bell,
524: {\it Mon. Not. Roy. Astron. Soc.} {\bf 182}, 147 (1978)
525:
526: \bibitem{bo78}
527: R. D. Blandford, J. P. Ostriker,
528: {\it Astrophys. J.} {\bf 221}, L29 (1978)
529:
530: \bibitem{biermann50}
531: L. Biermann,
532: {\it Z. Naturforsch} {\bf 5a}, 65 (1950)
533:
534: \bibitem{weibel59}
535: E. S. Weibel,
536: {\it Phys. Rev. Lett.} {\bf 2}, 83 (1959)
537:
538: \bibitem{msk06}
539: M. V. Medvedev, L. O. Silva, M. Kamionkowski,
540: {\it Astrophys. J.} {\bf 642}, L1 (2006)
541:
542: \bibitem{binn74}
543: J. Binney,
544: {\it Mon. Not. Roy. Astron. Soc.} {\bf 168}, 73 (1974)
545:
546: \bibitem{dw00}
547: G. Davies, L. M. Widrow,
548: {\it Astrophys. J.} {\bf 540}, 755 (2000)
549:
550: \bibitem{va99}
551: H. J. V\"olk, A. M. Atoyan,
552: {\it Astropart. Phys.} {\bf 11}, 73 (1999).
553:
554: \bibitem{kdlc01}
555: P. P. Kronberg, Q. W. Dufton, H. Li, S. A. Colgate,
556: {\it Astrophys. J.} {\bf 560}, 178 (2001)
557:
558: \bibitem{ssh06}
559: K. Subramanian, K., A. Shukurov, N. E. L. Haugen,
560: {\it Mon. Not. Roy. Astron. Soc.} {\bf 366}, 1437 (2006).
561:
562: \bibitem{rokc93}
563: D. Ryu, J. P. Ostriker, H. Kang, R. Cen,
564: {\it Astrophys. J.} {\bf 414}, 1 (1993)
565:
566: \bibitem{kz07}
567: R. M. Kulsrud, E. G. Zweibel,
568: {\it Rep. Prog. Phys.} {\bf 71}, 046901 (2008)
569:
570: \bibitem{cv00}
571: J. Cho, E. T. Vishniac,
572: {\it Astrophys. J.} {\bf 538}, 217 (2000)
573:
574: \bibitem{gbf04}
575: B. M. Gaensler, R. Beck, L. Feretti,
576: {\it New Astron. Rev.} {\bf 48}, 1003 (2004)
577:
578: \bibitem{dkrc08}
579: S. Das, H. Kang, D. Ryu, J. Cho
580: {\it Astrophys. J.} {\bf 681}, in press (2008)% (astro-ph/arXiv: 0801.0371)
581:
582: \end{thebibliography}
583:
584: \begin{scilastnote}
585: \item This work was supported by KICOS through grant
586: K20702020016-07E0200-01610 provided by MOST.
587: The work of D.R. was also supported by KOSEF through the grant of
588: basic research program R01-2007-000-20196-0.
589: The work of H.K. and S.D. was also supported by KOSEF through
590: ARCSEC.
591: The work of J.C. was also supported by KRF through grant
592: KRF-2006-331-C00136 funded by the Korean Government.
593: \end{scilastnote}
594:
595: \clearpage
596:
597: \centerline{\large \bf Supporting Online Material}
598:
599: \vspace{3mm}
600:
601: \noindent{\large \bf Materials and Methods}
602:
603: \vspace{3mm}
604:
605: \noindent{\bf S1. Simulation of large scale structure formation}
606: \hfill\break
607: The simulation used to estimate the vorticity in the intergalactic
608: medium (IGM) is for a $\Lambda$CDM universe with the following values
609: of cosmological parameters:
610: $\Omega_{BM}=0.043$, $\Omega_{DM}=0.227$, $\Omega_{\Lambda}=0.73$,
611: $h \equiv H_0$/(100 km/s/Mpc) = 0.7, and $\sigma_8 = 0.8$.
612: It was performed using a PM/Eulerian hydrodynamic code for the formation
613: of large scale structure (LSS) ({\it S1}), in a cubic region of comoving
614: volume $(100 \Mpc)^3$ with $1024^3$ grid zones for gas and gravity and
615: $512^3$ particles for dark matter, allowing a spatial resolution of
616: $\Delta l = 97.7$ kpc.
617: The simulation is adiabatic in the sense that it does not include
618: radiative cooling, galaxy/star formation, feedbacks from
619: galaxies/stars, and reionization of the IGM.
620: Although these processes play important roles in determining the
621: physical conditions mainly in cluster cores and voids, it was
622: shown that they do not affect significantly the global properties
623: and dissipations of cosmological shocks ({\it S2, S3}).
624:
625: \vspace{3mm}
626:
627: \noindent{\bf S2. Phases of the IGM}
628: \hfill\break
629: The intergalactic gas is heated mainly by cosmological shocks
630: ({\it S4, S5}).
631: It was suggested that the IGM can be classified roughly into four
632: phases according to gas temperature:
633: the hot gas with $T > 10^7$K mostly inside and around clusters/groups,
634: the warm-hot intergalactic medium (WHIM) with $T = 10^5 - 10^7$K
635: found mostly in filaments, the low temperature WHIM with
636: $T = 10^4 - 10^5$K distributed mostly as sheetlike structures,
637: and the diffuse gas with $T < 10^4$K residing mostly in voids
638: ({\it S5, S6}).
639: In particular, the intracluster medium (ICM) refers to the hottest
640: phase of the IGM inside clusters that is observable in x-ray.
641:
642: \vspace{3mm}
643:
644: \noindent{\bf S3. Vorticity generation}
645: \hfill\break
646: Vorticity can be generated directly at curved shocks and by the
647: baroclinity of flows.
648: For uniform upstream flow, the vorticity produced behind curved
649: shock surface is
650: $$
651: {\vec \omega}_{\rm cs} = \frac{(\rho_2 - \rho_1)^2}{\rho_1 \rho_2}
652: K {\vec U}_1 \times {\hat n},
653: \eqno{(S1)}
654: $$
655: where $\rho_1$ and $\rho_2$ are the upstream and downstream gas
656: densities, respectively, ${\vec U}_1$ is the upstream flow velocity
657: in the shock rest frame, $K$ is the curvature tensor of the shock
658: surface, and ${\hat n}$ is the unit vector normal to the surface
659: ({\it S7}).
660: If isopycnic surfaces (constant density surfaces) do not coincide
661: with isobaric surfaces, vorticity is generated with the rate given
662: by
663: $$
664: {\dot{\vec \omega}}_{\rm bc} = \frac{1}{\rho^2}
665: {\vec\nabla}\rho\times{\vec\nabla}p.
666: \eqno{(S2)}
667: $$
668:
669: Rough estimates of the vorticity generations by the two processes
670: in the simulation of LSS formation are made as follows.
671: Considering the generation is a random walk process,
672: $$
673: ({\dot\omega}\Delta t)_{\rm rms}
674: \left(\frac{t_{\rm age}(z)}{\Delta t}\right)^{1/2}
675: \eqno{(S3)}
676: $$
677: is calculated.
678: Here $\Delta t$ is the duration of coherent generation, and so
679: $t_{\rm age}(z)/\Delta t$ is the number of random walks during
680: the age of the universe at redshift $z$.
681: For the direct generation at cosmological shocks,
682: $$
683: {\dot\omega}_{\rm cs} \sim \frac{(\rho_2 - \rho_1)^2}{\rho_1 \rho_2}
684: \frac{U_{n1} U_{n2}}{R}\frac{g}{\Delta l}
685: \eqno{(S4)}
686: $$
687: is used.
688: Here $R$ is the curvature radius of the shocks, $U_{n1}$ and $U_{n2}$
689: are the flow speeds normal to the shocks in upstream and downstream,
690: $\Delta l$ is the grid size, and
691: $g$ takes account of all other factors such
692: as geometric ones.
693: For the baroclinic vorticity generation,
694: ${\dot\omega}_{\rm bc}$ from Eq. ({\it S2}) is used.
695: In the estimations, the distribution and property of shocks and flows
696: from the simulation are used.
697: Also $\Delta t$ is taken to be $10^8$ yrs, which is a typical
698: dynamical time, e.g., the sound crossing time in cluster cores.
699: The comoving value of $\left<R\right> \sim 1 \Mpc /(1+z)$, and
700: $\left<g\right> \sim 1$ are taken.
701: Note that $1 \Mpc$ is the typical scale of nonlinear structures
702: at present.
703: The factor $1/(1+z)$ is included because the average curvature radius
704: of cosmological shocks is smaller in the past even in the comoving
705: frame ({\it S8}).
706: In Fig. 2 the quantities normalized with $1/t_{\rm age}(z)$ are shown.
707:
708: \vspace{3mm}
709:
710: \noindent{\bf S4. Decomposition of flow velocity}
711: \hfill\break
712: The flow velocity can be decomposed into
713: $$
714: {\vec v} = {\vec v}_{\rm div} + {\vec v}_{\rm curl} +{\vec v}_{\rm unif},
715: \eqno{(S5)}
716: $$
717: where the divergence and curl components are defined as
718: ${\vec\nabla}\cdot{\vec v}_{\rm div} \equiv {\vec\nabla}\cdot{\vec v}$ and
719: ${\vec\nabla}\times{\vec v}_{\rm curl} \equiv {\vec\nabla}\times{\vec v}$,
720: respectively.
721: That is, ${\vec v}_{\rm div}$ is associated to compressional motions,
722: while ${\vec v}_{\rm curl}$ to incompressible shear motions.
723: Here ${\vec v}_{\rm unif}$ is the component uniform across the
724: computational box, whose magnitude is much smaller than the other
725: two components.
726: The decomposition is calculated exactly in Fourier space.
727: We note with the above decomposition, locally
728: $$
729: {\vec v}_{\rm div}\cdot{\vec v}_{\rm curl} \ne 0
730: ~~~~~~~~{\rm so}~~~~~~~~
731: \frac{1}{2} v^2 \ne
732: \frac{1}{2} \left(v_{\rm div}^2+ v_{\rm curl}^2 \right).
733: \eqno{(S6)}
734: $$
735: However, globally
736: $$
737: \int_{\rm box} {\vec v}_{\rm div}\cdot{\vec v}_{\rm curl} d^3{\vec x} = 0
738: ~~~~~~~~{\rm so}~~~~~~~~
739: \int_{\rm box} \frac{1}{2} v^2 d^3{\vec x} = \int_{\rm box}
740: \frac{1}{2} \left(v_{\rm div}^2+ v_{\rm curl}^2 \right) d^3{\vec x}.
741: \eqno{(S7)}
742: $$
743:
744: \vspace{3mm}
745:
746: \noindent{\bf S5. Seed magnetic fields}
747: \hfill\break
748: A number of mechanisms that can create seeds field in the early
749: universe have been suggested.
750: Besides various inflationary and string theory mechanisms, the most
751: popular astrophysical mechanism is the Biermann battery ({\it S9}).
752: It was discussed in the context of cosmological shocks ({\it S10})
753: and ionization fronts ({\it S11}), and shown to build up
754: weak fields of strength up to $10^{-20}$ G by $z \sim$ a few.
755:
756: At cosmological shocks, in addition, Weibel instability can
757: operate and produce magnetic fields up to the level of
758: $\varepsilon_B \sim 10^{-3} \varepsilon_{\rm sh}$ ({\it S12}),
759: and streaming cosmic rays accelerated by the shocks can amplify
760: weak upstream magnetic fields up to the level of
761: $\varepsilon_B \sim (1/2)(U_1/c) \varepsilon_{\rm CRs}$
762: via non-resonant growing mode ({\it S13}).
763: Here $\varepsilon_{\rm sh}$ is the energy density of upstream flow,
764: and $\varepsilon_B$ and $\varepsilon_{\rm CRs}$ are the
765: energy densities of downstream magnetic fields and CRs.
766: With $U_{1}/c \sim 10^{-3}$ for cosmological shocks
767: ({\it S2}, {\it S3}), these processes can potentially produce stronger
768: seed fields, although the coherence length of the resulting fields
769: is expected to very small and microscopic.
770:
771: \vspace{3mm}
772:
773: \noindent{\bf S6. Simulation of MHD turbulence}
774: \hfill\break
775: First we note there are at least two distinct types of
776: magnetohydrodynamic (MHD) turbulence:
777: one with strong regular fields and the other with weak/zero regular
778: fields.
779: The former has been described successfully with the nonlinear
780: interactions of Alfv\'en waves ({\it S14}).
781: The latter is mainly hydrodynamic in large scales,
782: but is modified by dynamically
783: important magnetic fields in small scales ({\it S15}, {\it S16}).
784: It is the latter that is relevant to this work.
785:
786: Incompressible, driven turbulence was simulated initially with a
787: very weak regular field.
788: A pseudo-spectral code ({\it S15}) was used, employing hyper-viscosity
789: and hyper-resistivity with the Prandtl number of unity.
790: The advantages of this numerical approach include virtually zero
791: intrinsic numerical viscosity and resistivity, and the maximized
792: inertial range.
793: The simulation was performed with a numerical resolution of $256^3$
794: collocation points.
795: The turbulence was driven at the scale of $L_{\rm driving} \sim
796: (1/2)L_{\rm box}$, where $L_{\rm box}$ is the computational box size.
797: The driving strength was set so that the total turbulence energy
798: becomes $\varepsilon_{\rm turb} \equiv \varepsilon_{\rm kin} +
799: \varepsilon_B \sim 1$ at saturation.
800: Initially $\varepsilon_B=10^{-6}$, but the evolution is not sensitive
801: to the initial field strength as long as it is sufficiently weak
802: ({\it S16}).
803: The left panel of Fig. S2 shows the time evolution of kinetic and
804: magnetic energies.
805: Here, the eddy turnover time is defined with the vorticity at driving
806: scale at saturation, $t_{\rm eddy} \equiv 1/\omega_{\rm driving}$.
807:
808: We made the incompressible simulation, because compressible simulations
809: need much higher resolution to achieve the same growth rate and
810: magnetic field strength at saturation.
811: In addition, due to large numerical dissipation, controlling
812: viscosity and resistivity is not trivial in compressible simulations.
813: Nevertheless, in order to show the general behavior of turbulence
814: dynamo seen in our incompressible simulation remains the same in
815: compressible regime, we performed compressible simulations of driven
816: MHD turbulence, using a code based on the third-order essentially
817: non-oscillatory (ENO) upwind scheme ({\it S17}).
818: This is one of the numerical schemes for MHD with least numerical
819: dissipation.
820: Different numerical resolutions of $8^3$ to $216^3$ grid zones
821: were used.
822: But otherwise, the initial conditions, the driving of turbulence,
823: and hyper-viscosity and hyper-resistivity with the Prandtl number
824: of unity were the same as in the incompressible simulation.
825: We only considered the case with the Mach number $M_s \sim 1$, which
826: is most relevant to this work.
827: We note that turbulence dynamo can be suppressed
828: in highly supersonic flows ({\it S18}).
829:
830: Fig. S5 compares the growth rate (left panel) and the magnetic
831: energy at saturation (right panel) from compressible simulations
832: with those from incompressible simulation.
833: Two points are clear from the figure: (1) the growth rate is
834: slower in compressible simulations, but the pattern follows that
835: in incompressible turbulence,
836: and (2) compressible simulation would need more than $1000^3$
837: grid zones to achieve the saturated magnetic field strength of
838: incompressible simulation.
839: It also demonstrates why it is impossible to reproduce the full
840: development of MHD turbulence in the simulations of LSS formation
841: with the currently available computational resources.
842:
843: \vspace{3mm}
844:
845: \noindent{\bf S7. Convergence of our results}
846: \hfill\break
847: In order to test the numerical convergence of our estimation for
848: the strength of the intergalactic magnetic fields (IGMFs), we
849: repeated the same analysis for the simulations of LSS formation
850: with lower numerical resolutions of $64^3$ to $512^3$ grid zones.
851: Except the resolution, they are the same simulations as that in
852: \S S1, in the sense that they have the same box size, the same
853: realization of initial conditions, and the same physics included.
854: Fig. S6 shows the volume-averaged magnetic field strength with
855: different resolutions.
856: The averaged strength around clusters/groups with $T > 10^7$K is
857: well converged at the resolution of $1024^3$.
858: On the other hand the averaged strength in filaments with
859: $10^5 < T < 10^7$K would be underestimated.
860: The converged value may be $\sim 2 - 3$ times larger than our
861: estimate with the resolution of $1024^3$.
862:
863: \vspace{3mm}
864:
865: \noindent{\bf S8. Coherence length}
866: \hfill\break
867: The consideration of the coherence length of turbulent magnetic
868: fields raises the issues of the scale of energy equipartition
869: and the magnetic field structure at the scale.
870: The right panel of Fig. S2 shows the power spectra for flow
871: velocity and magnetic fields at a time of saturation from the
872: simulation in \S S6.
873: The energy equipartition occurs at $\sim (1/2.5) L_{\rm driving}$,
874: which is close to the scale of the magnetic energy peak,
875: $\sim (1/3) L_{\rm driving}$.
876: The kinetic energy peak occurs at the the driving scale,
877: $L_{\rm driving}$.
878: Above the equipartition scale the flow structure is nearly isotropic,
879: while below it both the flow and magnetic field structures are
880: anisotropic and the eddies are stretched along the local magnetic
881: field lines ({\it S15}).
882: On the other hand, the magnetic field has most power at
883: $\sim (1/6) L_{\rm driving}$.
884: So we may argue that in the fully developed stage of the MHD
885: turbulence considered here, the coherence length of magnetic fields
886: would be several times smaller than the driving scale, the scale
887: of dominant eddies.
888:
889: \vspace{3mm}
890:
891: \noindent{\bf Additional Figures Used in Manuscript}
892:
893: \vspace{3mm}
894:
895: \noindent
896: Fig. S1 shows the correlation between the magnitude of vorticity
897: and the gas density (left panel) and the strength of magnetic fields
898: and the gas density (right panel) at present.
899:
900: \vspace{3mm}
901:
902: \noindent
903: Fig. S3 shows the spatial distribution of our model IGMFs around
904: a cluster complex and along a filament.
905: In the left image, in addition to the main cluster of
906: $T_x \approx 3.3$ keV, a group has $B \sim$ a few $\mu$G.
907: They are surrounded by a broad region of $B \sim 10\nG$.
908: In the right image, groups with $B \sim \mG$ are distributed
909: along the filament with $B \sim 10\nG$.
910:
911: \vspace{3mm}
912:
913: \noindent
914: Fig. S4 shows the power spectra for the flow velocity and its curl
915: and divergence components (\S S4) at present from the simulation of
916: LSS formation in \S S1.
917: At long wavelengths, the amplitude of perturbations are small, so
918: that the linear theory applies.
919: That is, $P_{\rm curl}(k) \rightarrow 0$ as $k \rightarrow 0$,
920: while $P_{\rm div}(k)$ follows the analytic theory expectation,
921: $P_{\rm div}(k) \sim k^{-1}$.
922: For wavelengths smaller than a few $\Mpc$, nonlinearities dominate,
923: and we see that $P_{\rm curl}(k) \ga P_{\rm div}(k)$.
924: $P_{\rm curl}(k)$ peaks at $\sim 5\Mpc$, and for values of $k$ somewhat
925: larger than the peak wavenumber, the spectrum follows a power law of
926: $k^{-5/3}$, the Kolmogorov spectrum.
927: $P_{\rm curl}(k)$ has most power at $\sim 2\Mpc$, that indicates the
928: typical scale of nonlinear structures in the simulation.
929:
930: \clearpage
931:
932: \begin{figure}
933: \vspace{-4cm}
934: \centerline{\epsfxsize=16cm\epsfbox{sf1.eps}}
935: \vspace{-2.5cm}
936: {Fig. S1. Volume fraction with given gas density and vorticity
937: magnitude (left panel) and gas density and magnetic field strength
938: (right panel) at present.}
939: \end{figure}
940:
941: \clearpage
942:
943: \begin{figure}
944: \vspace{-4cm}
945: \centerline{\epsfxsize=16cm\epsfbox{sf2.eps}}
946: \vspace{-2.5cm}
947: {Fig. S2. Left panel: Time evolution of kinetic and magnetic
948: energies in a three-dimensional, incompressible simulation of driven
949: MHD turbulence with very weak initial magnetic field.
950: The green lines show our fitting for the growth and saturation of
951: magnetic energy.
952: Right panel: Power spectra for flow velocity, $P_v$, and magnetic
953: fields, $P_B$, at a time of saturation.
954: Two straight lines of slopes $-5/3$ and $-1$ are also drawn for
955: comparison.}
956: \end{figure}
957:
958: \clearpage
959:
960: \begin{figure}
961: \vspace{0cm}
962: \centerline{\epsfxsize=16cm\epsfbox{sf3.eps}}
963: \vspace{0.5cm}
964: {Fig. S3. Volume rendering images showing logarithmically scaled
965: magnetic field strength at $z=0$ in a volume of $(25 \Mpc)^3$ around
966: the same cluster complex as shown in Fig. 1 (left panel) and in a
967: volume of $25\times15.6\times6.25\ (h^{-1}{\rm Mpc})^3$ which includes
968: a number of groups along a filament (right panel).
969: As in Fig. 4, color codes the magnetic field strength from $0.1 \nG$
970: (yellow) to $10 \mG$ (magenta).
971: Clusters and groups are shown with magenta and blue while filaments
972: with green.}
973: \end{figure}
974:
975: \clearpage
976:
977: \begin{figure}
978: \vspace{-4cm}
979: \centerline{\epsfxsize=16cm\epsfbox{sf4.eps}}
980: \vspace{-2.5cm}
981: {Fig. S4. Power spectra, $\int P_v dk = \left<(1/2) v^2 \right>$,
982: for flow velocity and its curl and divergence components at present
983: in the simulation of LSS formation.
984: Two straight lines of slopes $-5/3$ and $-1$ are also drawn for
985: comparison.}
986: \end{figure}
987:
988: \clearpage
989:
990: \begin{figure}
991: \vspace{-4cm}
992: \centerline{\epsfxsize=16cm\epsfbox{sf5.eps}}
993: \vspace{-2.5cm}
994: {Fig. S5. Left panel: Time evolution of magnetic energy in
995: three-dimensional, compressible simulations of driven MHD
996: turbulence with very weak initial magnetic field.
997: For comparison, the evolution in incompressible simulation shown
998: in Fig. S3 is also plotted.
999: Right panel: Magnetic energy at saturation in compressible
1000: simulations with different resolutions of $N_{\rm cell}^3$.
1001: The value in incompressible simulation is marked with solid line.}
1002: \end{figure}
1003:
1004: \clearpage
1005:
1006: \begin{figure}
1007: \vspace{-4cm}
1008: \centerline{\epsfxsize=16cm\epsfbox{sf6.eps}}
1009: \vspace{-2.5cm}
1010: {Fig. S6. Volume-averaged magnetic field strength at present for
1011: four temperature phases of the IGM and for all the gas from structure
1012: formation simulations with different resolutions of $N_{\rm cell}^3$.}
1013: \end{figure}
1014:
1015: \clearpage
1016:
1017: \noindent{\bf References}
1018:
1019: \begin{enumerate}
1020:
1021: \item[S1.]
1022: D. Ryu, J. P. Ostriker, H. Kang, R. Cen,
1023: {\it Astrophys. J.} {\bf 414}, 1 (1993)
1024:
1025: \item[S2.]
1026: D. Ryu, H. Kang, E. Hallman, T. W. Jones,
1027: {\it Astrophys. J.} {\bf 593}, 599 (2003)
1028:
1029: \item[S3.]
1030: H. Kang, D. Ryu, R. Cen, J. P. Ostriker,
1031: {\it Astrophys. J.} {\bf 669}, 729 (2007)
1032:
1033: \item[S4.]
1034: R. Cen, J. P. Ostriker,
1035: {\it Astrophys. J.} {\bf 514}, 1 (1999)
1036:
1037: \item[S5.]
1038: R. Dav\'e \etal~
1039: {\it Astrophys. J.} {\bf 552}, 473 (2001)
1040:
1041: \item[S6.]
1042: H. Kang, D. Ryu, R. Cen, D. Song,
1043: {\it Astrophys. J.} {\bf 620}, 21 (2005)
1044:
1045: \item[S7.]
1046: W. D. Hayes,
1047: {\it J. Fluid Mech.} {\bf 2}, {595} (1957)
1048:
1049: \item[S8.]
1050: D. Ryu, H. Kang
1051: {\it J. Korean Astron. Soc.} {\bf 37}, 477 (2004)
1052:
1053: \item[S9.]
1054: L. Biermann,
1055: {\it Z. Naturforsch} {\bf 5a}, 65 (1950)
1056:
1057: \item[S10.]
1058: R. M. Kulsrud, R. Cen, J. P. Ostriker, D. Ryu,
1059: {\it Astrophys. J.} {\bf 480}, 481 (1997)
1060:
1061: \item[S11.]
1062: N. Y. Gnedin, A. Ferrara, E. G. Zweibel,
1063: {\it Astrophys. J.} {\bf 539}, 505 (2000)
1064:
1065: \item[S12.]
1066: M. V. Medvedev, L. O. Silva, M. Kamionkowski,
1067: {\it Astrophys. J.} {\bf 642}, L1 (2006)
1068:
1069: \item[S13.]
1070: A. R. Bell,
1071: {\it Mon. Not. Roy. Astron. Soc.} {\bf 353}, 550 (2004)
1072:
1073: \item[S14.]
1074: P. Goldreich, S. Sridhar,
1075: {\it Astrophys. J.} {\bf 438}, 763 (1995)
1076:
1077: \item[S15.]
1078: J. Cho, E. T. Vishniac,
1079: {\it Astrophys. J.} {\bf 538}, 217 (2000)
1080:
1081: \item[S16.]
1082: J. Cho, E. T. Vishniac, A.Beresnyak, A. Lazarian, D. Ryu,
1083: {\it Astrophys. J.} submitted (2008)
1084:
1085: \item[S17.]
1086: J. Cho, A. Lazarian,
1087: {\it Phys. Rev. Lett.} {\bf 88}, 245001 (2002)
1088:
1089: \item[S18.]
1090: N. E. L. Haugen, A. Brandenburg, A. J. Mee,
1091: {\it Mon. Not. Roy. Astron. Soc.} {\bf 353}, 947 (2004)
1092:
1093: \end{enumerate}
1094:
1095: \end{document}
1096: