1: \documentclass[runningheads]{svmult}

2: %

3: \usepackage{makeidx}   % allows index generation

4: %\usepackage{graphics}  % standard LaTeX graphics tool

5: %                       % for including eps-figure files

6: \usepackage{subeqnar}  % subnumbers individual equations

7: %                       % within an array

8: %\usepackage{multicol}  % used for the two-column index

9: %%\usepackage{cropmark} % cropmarks for pages without

10: %                       % pagenumbers - only needed when manuscript

11: %                       % is printed from paper and not from data

12: \usepackage{physprbb}  % centered layout of diverse elements, etc.

13: \makeindex             % used for the subject index

14: %                       % please use the style sprmidx.sty with

15: %                       % your makeindex program

16: %

17: %%%upright Greek letters (example below: upright ``mu'')

18: \newcommand{\greeksym}[1]{{\usefont{U}{psy}{m}{n}#1}}

19: \newcommand{\umu}{\mbox{\greeksym{m}}}

20: \newcommand{\udelta}{\mbox{\greeksym{d}}} 

21: \newcommand{\ga}{>}

22: \newcommand{\la}{<} 

23: \newcommand{\msun}{M_\odot}

24: \newcommand{\lya}{Ly$\alpha$~} 

25: \newcommand{\kel}{~{\rm K}}

26: \newcommand{\arcdeg}{^\circ} 

27: \newcommand{\uPi}{\mbox{\greeksym{P}}}

28: 

29: \begin{document}

30: 

31: \title*{Novel Ways to Probe the Universe with Gamma-Ray Bursts and Quasars}

32: 

33: \toctitle{Novel Ways to Probe the Universe with Gamma-Ray Bursts and

34: Quasars} 

35: \titlerunning{Novel Ways to Probe the Universe with Gamma-Ray

36: Bursts and Quasars}

37: 

38: 

39: \author{Abraham Loeb}

40: \authorrunning{Avi Loeb}

41: \institute{Astronomy Department, Harvard University\\

42: Cambridge, MA 02138, USA}

43: 

44: \maketitle              % typesets the title of the contribution

45: 

46: \begin{abstract}

47: 

48: I consider novel ways by which Gamma-Ray Bursts (GRBs) and quasars can be

49: used to probe the universe. Clues about how and when was the intergalactic

50: medium ionized can be read off the UV emission spectrum of GRB explosions

51: from the first generation of stars. The existence of intergalactic and

52: galactic stars can be inferred from their gravitational microlensing effect

53: on GRB afterglows.  Prior to reionization, quasars should be surrounded by

54: a halo of scattered Ly$\alpha$ radiation which probes the neutral

55: intergalactic medium (IGM) around them. The situation is analogous to the

56: appearance of a halo of scattered light around a street lamp which is

57: embedded in a dense fog. Outflows from quasars magnetize the IGM at all

58: redshifts. As a result, the shocks produced by converging flows during the

59: formation of large scale structure in the IGM, accelerate electrons to

60: relativistic energies and become visible in the radio regime through their

61: synchrotron emission and in the $\gamma$--ray regime through their

62: inverse--Compton scattering of the microwave background photons. During

63: transient episodes of strong mergers, X-ray clusters should therefore

64: appear as extended radio or $\gamma$--ray sources on the sky.

65: 

66: \end{abstract}

67: 

68: \section{Novel Cosmological Studies with Gamma-Ray Bursts}

69: 

70: \subsection{Preface}

71: 

72: Since their discovery four decades ago, quasars have been used as powerful

73: lighthouses which probe the intervening universe out to high redshifts,

74: $z\sim 6$ \cite{Hartwick,Fan}.  The spectra of almost all quasars show

75: strong emission lines of metals, indicating super-solar enrichment of the

76: emitting gas \cite{Hamann}. This implies that at least in the cores of

77: galaxies, formation of massive stars and their evolution to supernovae

78: preceded the observed quasar activity.  If Gamma-Ray Bursts (GRBs)

79: originate from the remnants of massive stars (such as neutron stars or

80: black holes), as seems likely based on recent estimates of their energy

81: output \cite{WKF,Freedman,Frail}, then they should exist at least out to

82: the same redshift as quasars.  Although GRBs are transient events, their

83: peak optical-UV flux can be as bright as that of quasars. Hence, GRBs

84: promise to be as useful as quasars in probing the high--redshift universe.

85: 

86: Not much is known observationally about the universe in the redshift

87: interval $z=6$--$30$, when the first generation of galaxies condensed out

88: of the primordial gas left over from the Big Bang (see reviews

89: \cite{Barkana} and \cite{Loeb_Barkana}). Observations of the microwave

90: background anisotropies indicate that the cosmic gas became neutral at

91: $z\sim 1000$ and remained so at least down to $z\sim 30$ (see,

92: e.g. \cite{Wang}).  On the other hand, the existence of transmitted flux

93: shortward of the Ly$\alpha$ resonance in the spectrum of the

94: highest-redshift quasars and galaxies (see, for example,

95: Figure~\ref{fig1}), indicates that the intergalactic medium was reionized

96: to a level better than 99.9999\% by a redshift $z\sim 6$.  This follows

97: from the fact that the Ly$\alpha$ optical depth of the intergalactic medium

98: at high-redshifts ($z\gg1$) is \cite{Gunn},

99: \begin{equation}

100: \tau_{\alpha}={\pi e^2 f_\alpha \lambda_\alpha n_{HI}(z) \over m_e

101: cH(z)} \approx 6.15\times 10^5 x_{HI} \left({\Omega_bh\over

102: 0.03}\right)\left({\Omega_m\over 0.3}\right)^{-1/2} \left({1+z\over

103: 10}\right)^{3/2}, 

104: \label{eq:G-P}

105: \end{equation}

106: where $H\approx 100h~{\rm km~s^{-1}~Mpc^{-1}}\Omega_m^{1/2}(1+z)^{3/2}$ is

107: the Hubble parameter at the source redshift $z$, $f_\alpha=0.4162$ and

108: $\lambda_\alpha=1216$\AA\, are the oscillator strength and the wavelength

109: of the Ly$\alpha$ transition; $n_{HI}(z)$ is the average intergalactic

110: density of neutral hydrogen at the source redshift (assuming primordial

111: abundances); $\Omega_m$ and $\Omega_b$ are the present-day density

112: parameters of all matter and of baryons, respectively; and $x_{HI}$ is the

113: average fraction of neutral hydrogen. Modeling \cite{Fan} of the {\it

114: transmitted flux} in Figure~\ref{fig1} implies $\tau_{\alpha}<0.5$ or

115: $x_{HI}< 10^{-6}$, i.e., most of the low-density gas throughout the

116: universe is ionized to a level of 99.999\% at $z\la 6$. However, there are

117: some dark intervals in the spectrum which could be indicative of regions

118: with a higher neutral fraction \cite{Djorgovski}. In fact, Becker et

119: al. \cite{Becker} reported the detection of an extended ($\ga 300$\AA~

120: long) dark interval just shortward of the Ly$\alpha$ emission line in the

121: spectrum of the newly discovered quasar SDSS 1030+0524 at $z=6.28$.  The

122: suppression of the flux by a factor $\ga 150$ may indicate the first

123: detection of the Gunn-Peterson trough, although caution is warranted since

124: the inferred optical depth $\tau_\alpha\ga 5$ can be produced by a neutral

125: fraction as small as $X_{HI}\sim 2 \times 10^{-4}$ according to

126: equation~(\ref{eq:G-P}).

127: 

128: \begin{figure}[htbp]

129: \special{psfile=fig1.ps hoffset=-225 voffset=-415 hscale=100 vscale=100}

130: \vspace{2.8in}

131: \caption{Spectrum of the quasar SDSS1044-0125 at $z=5.73$ (Djorgovski et

132: al. 2001), originally discovered by the Sloan Digital Sky Survey (Fan et

133: al. 2000). }

134: \label{fig1}

135: \end{figure}

136: 

137: The question: {\it how and when was the universe reionized?}  defines a new

138: frontier in observational cosmology \cite{Loeb_Barkana}. The UV spectrum of

139: GRB afterglows can be used to probe the ionization and thermal state of the

140: intergalactic gas during the epoch of reionization, at redshifts $z\sim

141: 7$--10 \cite{Miralda_Escude}.  The stretching of the temporal evolution of

142: GRB lightcurves by the cosmological redshift factor $(1+z)$, makes it

143: easier for an observer to react in time and measure a spectrum of their

144: optical-UV emission at its peak.

145: 

146: Energy arguments suggest that reionization resulted from photoionization

147: and not from collisional ionization \cite{Tegmark,Furlanetto}.  The

148: corresponding sources of UV photons were either stars or quasars.  Recent

149: simulations of the first generation of stars that formed out of the

150: primordial metal--free gas indicate that these stars were likely massive

151: \cite{Bromm,Abel}. If GRBs result from compact stellar remnants, such as

152: black holes or neutron stars, then the fraction of all stars that lead to

153: GRBs may have been higher at early cosmic times. This, however, is true

154: only if the GRB phenomena is triggered on a time scale much shorter than

155: the age of the universe at the corresponding redshift, which for $z\gg 1$

156: is $\sim 5.4\times 10^8~{\rm yr}~(h/0.7)^{-1}(\Omega_m/0.3)^{-1/2}

157: [(1+z)/10]^{-3/2}$. 

158: %This condition may not hold, for example, for neutron

159: %star binaries with an excessively long coalescence time.

160: 

161: \subsection{Properties of High-Redshift GRB Afterglows}

162: 

163: Young (days to weeks old) GRBs outshine their host galaxies in the optical

164: regime. In the standard hierarchical model of galaxy formation, the

165: characteristic mass and hence optical luminosity of galaxies and quasars

166: declines with increasing redshift \cite{Haiman,Stern,Barkana}. Hence, GRBs

167: should become easier to observe than galaxies or quasars at increasing

168: redshift. Similarly to quasars, GRB afterglows possess broad-band spectra

169: which extend into the rest-frame UV and can probe the ionization state and

170: metallicity of the IGM out to the epoch when it was reionized at redshifts

171: $z\sim 7$--$10$ \cite{Loeb_Barkana}.  Lamb \& Reichart \cite{Lamb} have

172: extrapolated the observed $\gamma$-ray and afterglow spectra of known GRBs

173: to high redshifts and emphasized the important role that their detection

174: might play in probing the IGM. Simple scaling of the long-wavelength

175: spectra and temporal evolution of afterglows with redshift implies that at

176: a fixed time lag after the GRB trigger in the observer's frame, there is

177: only a mild change in the {\it observed} flux at infrared or radio

178: wavelengths as the GRB redshift increases. Ciardi \& Loeb \cite{Ciardi}

179: demonstrated this behavior using a detailed extrapolation of the GRB

180: fireball solution into the non-relativistic regime (see the 2$\mu$m curves

181: in Figure~\ref{fig2}). Despite the strong increase of the luminosity

182: distance with redshift, the observed flux for a given observed age is

183: almost independent of redshift in part because of the special spectrum of

184: GRB afterglows (see Figure~\ref{fig4}), but mainly because afterglows are

185: brighter at earlier times and a given observed time refers to an earlier

186: intrinsic time in the source frame as the source redshift increases. The

187: mild dependence of the long-wavelength ($\lambda_{\rm obs}>1\mu$m) flux on

188: redshift stands in contrast to other high-redshift sources such as galaxies

189: or quasars, which fade rapidly with increasing redshift

190: \cite{Haiman,Stern,Barkana}. Hence, GRBs provide exceptional lighthouses

191: for probing the universe at $z=6$--30, at the epoch when the first stars

192: had formed.

193: 

194: \begin{figure}[htbp]

195: \special{psfile=fig2.ps hoffset=15 voffset=-335 hscale=55 vscale=55}

196: \vspace{3.25in}

197: \caption{Theoretical expectation for the observed afterglow flux of a GRB

198: as a function of its redshift (from Ciardi \& Loeb 2000).  The curves refer

199: to an observed wavelength of $5000$ \AA (thin lines) and $2 \mu$m (thick

200: lines). Different line types refer to different observed times after the

201: GRB trigger, namely 1 hour (solid line), 1 day (dotted) and 10 days

202: (dashed). The $5000$\AA~ flux is strongly absorbed at $z>4.5$ by

203: intergalactic hydrogen. However, at infrared and radio wavelengths the

204: observed afterglow flux shows only a mild dependence on the source

205: redshift.}

206: \label{fig2}

207: \end{figure}

208: 

209: Assuming that the GRB rate is proportional to the star formation rate and

210: that the characteristic energy output of GRBs is $\sim 10^{52}~{\rm ergs}$,

211: Ciardi \& Loeb \cite{Ciardi} predicted that there are always $\sim 15$ GRBs

212: from redshifts $z> 5$ across the sky which are brighter than $\sim 100$ nJy

213: at an observed wavelength of $\sim 2\mu$m.  The infrared spectrum of these

214: sources could be taken with future telescopes such as the {\it Next

215: Generation Space Telescope} (planned for launch in 2009; see

216: http://ngst.gsfc.nasa.gov/), as a follow-up on their early X-ray

217: localization with the {\it Swift} satellite (planned for launch in 2003;

218: see http://swift.sonoma.edu/).

219: 

220: The redshifts of GRB afterglows can be estimated photometrically from

221: either the Lyman limit or Ly$\alpha$ troughs in their spectra. At low

222: redshifts, the question of whether the Lyman limit or Ly$\alpha$ trough

223: interpretation applies depends on the absorption properties of the host

224: galaxy. If the GRB originates from within the disk of a star--forming

225: galaxy, then the afterglow spectrum will likely show a damped Ly$\alpha$

226: trough. At $z> 6$ the Ly$\alpha$ trough would inevitably exist since the

227: intergalactic Ly$\alpha$ opacity is $> 90\%$ (see Figure 13 in

228: \cite{Stern}). Interestingly, an absorption feature in the afterglow

229: spectrum which is due to the neutral hydrogen within a molecular cloud or

230: the disk of the host galaxy, is likely to be time-dependent due to the

231: ionization caused by the UV illumination of the afterglow itself along the

232: line-of-sight \cite{Perna}.

233: 

234: So far, there have been two claims for high-redshift GRBs. Fruchter

235: \cite{Fruchter} argued that the photometry of GRB 980329 is consistent with

236: a Ly$\alpha$ trough due to IGM absorption at $z\sim 5$.  Anderson et

237: al. \cite{Anderson} inferred a redshift of $z=4.5$ for GRB 000131 based on

238: a crude optical spectrum that was taken by the VLT a few days after the GRB

239: trigger. {\it These cases emphasize the need for a coordinated observing

240: program that will alert 10-meter class telescopes to take a spectrum of an

241: afterglow about a day after the GRB trigger, based on a photometric

242: assessment (obtained with a smaller telecope using the Lyman limit or

243: Ly$\alpha$ troughs) that the GRB may have originated at a high redshift.}

244: 

245: In the following two subsections, I illustrate the usefulness of GRB

246: afterglows for cosmological studies through two examples.

247: 

248: \begin{figure}[htbp]

249: \special{psfile=fig3.ps hoffset=15 voffset=-360 hscale=48 vscale=48}

250: \vspace{4.2in}

251: \caption{ Sketch of the expected spectrum of a source at a redshift $z_{s}$

252: slightly above the reionization redshift $z_{\rm reion}$. The transmitted

253: flux due to HII bubbles in the pre-reionization era, and the Ly$\alpha$

254: forest in the post-reionization era, are exaggerated for illustration.  }

255: \label{fig3}

256: \end{figure}

257: 

258: \subsection{Probing the Reionization Epoch and Metallicity History of the IGM}

259: 

260: The UV spectra of GRB afterglows can be used to measure the evolution of

261: the neutral intergalactic hydrogen with redshift. Figure~\ref{fig3}

262: illustrates schematically the expected absorption just beyond the

263: reionization redshift. Resonant scattering suppresses the spectrum at all

264: wavelengths corresponding to the Ly$\alpha$ resonance prior to

265: reionization.  Since the Ly$\alpha$ cross-section is very large, any

266: transmitted flux prior to reionization reflects a large volume of ionized

267: hydrogen along the line-of-sight. If the GRB is located at a redshift

268: larger by $>18\%$ than the reionization redshift, then the Ly$\alpha$ and

269: the Ly$\beta$ troughs will overlap. Unlike quasars, GRBs do not ionize the

270: IGM around them; their limited energy supply $\sim 10^{52}~{\rm ergs}$

271: \cite{WKF,Freedman,Frail} can ionize only $\sim 4\times 10^5M_\odot$ of

272: neutral hydrogen within their host galaxy.

273: 

274: Quasar spectra indicate the existence of an IGM metallicity which is a

275: fraction of a percent of the solar value \cite{Ellison}.  The metals were

276: likely dispersed into the IGM through outflows from galaxies, driven by

277: either supernova or quasar winds \cite{Barkana,Furlanetto}. Detection of

278: metal absorption lines in the spectrum of GRB afterglows, produced either

279: in the IGM or the host galaxy of the GRB, can help unravel the evolution of

280: the IGM metallicity with redshift and its link to the evolution of

281: galaxies. Detection of X-ray absorption by intergalactic metals can be used

282: to establish the existence of the warm component of the IGM which has not

283: been observed so far \cite{Fiore,Hellsten,Perna}.

284: 

285: 

286: \subsection{Cosmological Microlensing of Gamma-Ray Bursts}

287: 

288: Loeb \& Perna \cite{Loeb_Perna} noted the coincidence between the angular

289: size of a solar-mass lens at a cosmological distance and the

290: micro-arcsecond size of the image of a GRB afterglow. They therefore

291: suggested that microlensing by stars can be used to resolve the

292: photospheres of GRB fireballs at cosmological distances.  (Alternative

293: methods, such as radio scintillations, only provide a constraint on the

294: radio afterglow image size \cite{Goodman,WKF} but do not reveal its

295: detailed surface brightness distribution, because of uncertainties in the

296: scattering properties of the Galactic interstellar medium.)

297: 

298: \begin{figure}[htbp]

299: \special{psfile=fig4.eps hoffset=0 voffset=-350 hscale=55 vscale=55}

300: \vspace{3.6in}

301: \caption{A typical broken power-law spectrum of a GRB afterglow at a

302: redshift $z=1$ (from Granot \& Loeb 2001). The observed flux density,

303: $F_\nu$, as a function of frequency, $\nu$, is shown by the boldface solid

304: line at an observed time $t_{\rm days}=1$ for an explosion with a total

305: energy output of $10^{52}~{\rm ergs}$ in a uniform interstellar medium

306: ($k=0$) with a hydrogen density of $1~{\rm cm^{-3}}$, and post-shock energy

307: fractions in accelerated electrons and magnetic field of $\epsilon_e=0.1$

308: and $\epsilon_B=0.03$, respectively.  The thin solid line shows the same

309: spectrum when it is microlensed by an intervening star with an impact

310: parameter equal to the Einstein angle and $R_0\equiv [\theta_s(1~{\rm

311: day})/\theta_{\rm E}]=1$. The insert shows the excess evolution of the

312: break frequencies $\nu_{\rm b}=\nu_a,~\nu_m$ and $\nu_c$ (normalized by

313: their unlensed values) due to microlensing. }

314: \label{fig4}

315: \end{figure}

316: 

317: The fireball of a GRB afterglow is predicted to appear on the sky as a ring

318: (in the optical band) or a disk (at low radio frequencies) which expands

319: laterally at a superluminal speed, $\sim \Gamma c$, where $\Gamma\gg1$ is

320: the Lorentz factor of the relativistic blast wave which emits the afterglow

321: radiation \cite{Waxman,Sari,Panaitescu,Granot}.  For a spherical explosion

322: into a constant density medium (such as the interstellar medium), the

323: physical radius of the afterglow image is of order the fireball radius over

324: $\Gamma$, or more precisely \cite{Granot}

325: \begin{equation}

326: R_{\rm s}= 3.9\times 10^{16}\left({E_{52}\over n_1}\right)^{1/8}

327: \left({t_{\rm days}\over 1+z}\right)^{5/8}~{\rm cm},

328: \label{eq:r_s}

329: \end{equation}

330: where $E_{52}$ is the hydrodynamic energy output of the GRB explosion in

331: units of $10^{52}~{\rm ergs}$, $n_1$ is the ambient medium density in units

332: of $1~{\rm cm^{-3}}$, and $t_{\rm days}$ is the observed time in days.  At

333: a cosmological redshift $z$, this radius of the GRB image occupies an angle

334: $\theta_{\rm s}=R_{\rm s}/D_A$, where $D_A(z)$ is the angular diameter

335: distance at the GRB redshift, $z$. For the typical cosmological distance,

336: $D_A\sim 10^{28}~{\rm cm}$, the angular size is of order a micro-arcsecond

337: ($\mu$as).  Coincidentally, this image size is comparable to the Einstein

338: angle of a solar mass lens at a cosmological distance,

339: \begin{equation}

340: \theta_{\rm E}=\left({4GM_{\rm lens}\over c^2 D}\right)^{1/2}= 1.6

341: \left({M_{\rm lens}\over 1 M_\odot}\right)^{1/2}\left({D\over 10^{28}~{\rm

342: cm}}\right)^{-1/2}~\mu{\rm as},

343: \label{eq:1}

344: \end{equation}

345: where $M_{\rm lens}$ is the lens mass, and $D\equiv {D_{\rm os}D_{\rm ol}/

346: D_{\rm ls}}$ is the ratio of the angular-diameter distances between the

347: observer and the source, the observer and the lens, and the lens and the

348: source \cite{Schneider}.  Loeb \& Perna \cite{Loeb_Perna} showed that

349: because the ring expands laterally faster than the speed of light, the

350: duration of the microlensing event is only a few days rather than tens of

351: years, as is the case for more typical astrophysical sources which move at

352: a few hundred ${\rm km~s^{-1}}$ or $\sim 10^{-3}c$.

353: 

354: The microlensing lightcurve goes through three phases: (i) constant

355: magnification at early times, when the source is much smaller than the

356: source-lens angular separation; (ii) peak magnification when the ring-like

357: image of the GRB first intersects the lens center on the sky; and (iii)

358: fading magnification as the source expands to larger radii.

359: 

360: \begin{figure}[htbp]

361: \special{psfile=fig5.eps hoffset=30 voffset=-350 hscale=45 vscale=45}

362: \vspace{4.8in}

363: \caption{The surface brightness, normalized by its average value, as a

364: function of the normalized radius, $r$, from the center of a GRB afterglow

365: image (where $r=0$ at the center and $r=1$ at the outer edge). The image

366: profile changes considerably between different power-law segments of the

367: afterglow spectrum, $F_{\nu}\propto\nu^{\beta}$ (see

368: Figure~\ref{fig4}). There is also a strong dependence on the power--law

369: index of the radial density profile of the external medium around the

370: source, $\rho\propto R^{-k}$ (taken from Granot \& Loeb 2001).  }

371: \label{fig5}

372: \end{figure}

373: 

374: Granot \& Loeb \cite{Granot_Loeb} calculated the radial surface brightness

375: profile (SBP) of the image of a Gamma-Ray-Burst (GRB) afterglow as a

376: function of frequency and ambient medium properties, and inferred the

377: corresponding magnification lightcurves due to microlensing by an

378: intervening star. The afterglow spectrum consists of several power-law

379: segments separated by breaks, as illustrated by Figure~\ref{fig4}. The

380: image profile changes considerably across each of the spectral breaks, as

381: shown in Figure~\ref{fig5}. It also depends on the power--law index, $k$,

382: of the radial density profile of the ambient medium into which the GRB

383: fireball propagates.  Gaudi \& Loeb \cite{Gaudi_Loeb} have shown that

384: intensive monitoring of a microlensed afterglow lightcurve can be used to

385: reconstruct the parameters of the fireball and its environment.  The

386: dependence of the afterglow image on frequency offers a fingerprint that

387: can be used to identify a microlensing event and distinguish it from

388: alternative interpretations.  It can also be used to constrain the

389: relativistic dynamics of the fireball and the properties of its gaseous

390: environment. At the highest frequencies, the divergence of the surface

391: brightness near the edge of the afterglow image ($r=1$ in

392: Figure~\ref{fig5}) depends on the thickness of the emitting layer behind

393: the relativistic shock front, which is affected by the length scale

394: required for particle acceleration and magnetic field amplification behind

395: the shock\cite{Medvedev,Gruzinov}.

396: 

397: Ioka \& Nakamura \cite{Ioka} considered the more complicated case where the

398: explosion is collimated and centered around the viewing axis. More general

399: orientations that violate circular symmetry need to be considered in the

400: future.

401: 

402: \begin{figure}[htbp]

403: \special{psfile=fig6.eps hoffset=25 voffset=-350 hscale=50 vscale=50}

404: \vspace{3.9in}

405: \caption{ $UBVRIJK$ photometry of GRB 000301C as a function of time in days

406: from the GRB trigger (from Garnavich et al. 2000; Gaudi et al.  2001).

407: Data points have been offset by the indicated amount for clarity.  The

408: dashed line is the best-fit smooth, double power-law lightcurve (with no

409: lensing), while the solid line is the overall best-fit microlensing model,

410: where the SBP has been determined from direct inversion.  The inset shows

411: the $R$-band data only.  The dotted line is the best-fit microlensing model

412: with theoretically calculated SBP, for $k=0$ and $\nu>\nu_c$.  }

413: \label{fig6}

414: \end{figure}

415: 

416: \subsubsection{GRB 000301C}

417: 

418: Garnavich, Loeb, \& Stanek \cite{Garnavich} have reported the possible

419: detection of a microlensing magnification feature in the optical-infrared

420: light curve of GRB 000301C (see Figure~\ref{fig6}). The achromatic

421: transient feature is well fitted by a microlensing event of a $0.5 M_\odot$

422: lens separated by an Einstein angle from the source center, and resembles

423: the prediction of Loeb \& Perna \cite{Loeb_Perna} for a ring-like source

424: image with a narrow fractional width ($\sim 10\%$). Alternative

425: interpretations relate the transient achromatic brightening to a higher

426: density clump into which the fireball propagates \cite{Berger}, or to a

427: refreshment of the decelerating shock either by a shell which catches up

428: with it from behind or by continuous energy injection from the source

429: \cite{Zhang}. However, the microlensing model has a smaller number of free

430: parameters. If with better data, a future event will show the generic

431: temporal and spectral characteristics of a microlensing event, then these

432: alternative interpretations will be much less viable.  A galaxy

433: 2$^{\prime\prime}$ from GRB~000301C might be the host of the stellar lens,

434: but current data provides only an upper-limit on its surface brightness at

435: the GRB position. The existence of an intervening galaxy increases the

436: probability for microlensing over that of a random line-of-sight.

437: 

438: \begin{figure}[htbp]

439: \special{psfile=fig7.eps hoffset=20 voffset=-350 hscale=50 vscale=50}

440: \vspace{3.9in}

441: \caption{Fitting GRB 000301C with different SBPs as a function of

442: normalized radius (taken from Gaudi et al. 2001).  The points are the SBPs

443: determined from direct inversion, with $1\sigma$ errors (solid) and

444: $3\sigma$ errors (dotted).  The curves are theoretically calculated SBPs

445: for various frequency regimes (see Figure~\ref{fig5}).  (a) Uniform

446: external medium, $k=0$.  (b) Stellar wind environment, $k=2$. The number of

447: degrees of freedom is 92 for the direct inversion points and 89 for the

448: curves.}

449: \label{fig7}

450: \end{figure}

451: 

452: Gaudi, Granot, \& Loeb \cite{GGL} have shown that direct inversion of the

453: observed light curve for GRB 000301C yields a surface brightness profile

454: (SBP) of the afterglow image which is strongly limb-brightened, as expected

455: theoretically (see Figure~\ref{fig7}).

456: 

457: Obviously, realistic lens systems could be more complicated due to the

458: external shear of the host galaxy or a binary companion.  Mao \& Loeb

459: \cite{Mao} calculated the magnification light curves in these cases, and

460: found that binary lenses may produce multiple peaks of magnification.  They

461: also demonstrated that {\it all} afterglows are likely to show variability

462: at the level of a few percent about a year following the explosion, due to

463: stars which are separated by tens of Einstein angles from the

464: line-of-sight.

465: 

466: {\it What is the probability for microlensing?} If the lenses are not

467: strongly clustered so that their cross-sections overlap on the sky, then

468: the probability for having an intervening lens star at a projected angular

469: separation $\theta$ from a source at a redshift $z\sim 2$ is~~$\sim 0.3

470: \Omega_\star (\theta/\theta_E)^2$ \cite{Press,Blaes,Nemiroff,Nemiroff2},

471: where $\Omega_\star$ is the cosmological density parameter of stars.  The

472: value of $\Omega_\star$ is bounded between the density of the luminous

473: stars in galaxies and the total baryonic density as inferred from Big Bang

474: nucleosynthesis, $7\times 10^{-3} < \Omega_\star< 5\times 10^{-2}$

475: \cite{Fukugita}. Hence, {\it all} GRB afterglows should show evidence for

476: events with $\theta \sim 30\theta_E$, for which microlensing provides a

477: small perturbation to the light curve\cite{Mao}. (This crude estimate

478: ignores the need to subtract those stars which are embedded in the dense

479: central regions of galaxies, where macrolensing dominates and the

480: microlensing optical depth is of order unity.) However, only one out of

481: roughly a hundred afterglows is expected to be strongly microlensed with an

482: impact parameter smaller than the Einstein angle. Indeed, Koopmans \&

483: Wambsganss \cite{Koopmans} have found that the `a posteriori' probability

484: for the observed microlensing event of GRB 000301C along a random

485: line-of-sight is between 0.7--2.7\% if 20-100\% of the dark matter is in

486: compact objects.

487: 

488: Microlensing events are rare but precious.  Detailed monitoring of a few

489: strong microlensing events among the hundreds of afterglows detected per

490: year by the forthcoming Swift satellite, could be used to constrain the

491: environment and the dynamics of relativistic GRB fireballs, as well as

492: their magnetic structure and particle acceleration process.

493: 

494: \section{Illumination of the Intergalactic Medium by Quasars}

495: 

496: \subsection{Ly$\alpha$ Halos}

497: 

498: The absorption trough created in the spectrum of a distant sources due to

499: neutral hydrogen diminishes down to undetectable flux levels as soon as the

500: mean neutral fraction of hydrogen is larger than $\sim 10^{-4}$ (see

501: equation~\ref{eq:G-P}). Once the transmitted flux reaches very low levels,

502: it is difficult to infer whether the neutral fraction of hydrogen is as low

503: as 0.01\% or as high as 100\%.  A novel way to study the reionization phase

504: transition relies on a direct detection of intergalactic neutral

505: hydrogen. Loeb \& Rybicki \cite{Loeb_Rybicki,Rybicki} have shown that the

506: neutral gas prior to reionization can be probed through narrow-band imaging

507: of embedded Ly$\alpha$ sources. The physical situation is analogous to the

508: appearance of a halo of scattered light around a street lamp which is

509: embedded in a dense fog. The ``street lamp'' in this metaphor is a

510: high--redshift quasar which emits Ly$\alpha$ photons into the surrounding

511: neutral hydrogen gas.  The IGM scatters these photons and acts as a fog.

512: 

513: The radiation of the first galaxies is strongly absorbed shortward of their

514: restframe Ly$\alpha$ wavelength by neutral hydrogen in the intervening IGM.

515: However, the Ly$\alpha$ photons emitted by these sources are not eliminated

516: but rather scatter until they redshift out of resonance and escape due to

517: the Hubble expansion of the surrounding intergalactic hydrogen (see

518: Fig. 8).  Detection of the diffuse Ly$\alpha$ halos around high redshift

519: sources would provide a unique tool for probing the neutral IGM before the

520: epoch of reionization. Loeb \& Rybicki \cite{Loeb_Rybicki,Rybicki} explored

521: the above effect for a uniform, fully-neutral IGM in a pure Hubble flow.

522: It is important to extend this analysis to more realistic cases of sources

523: embedded in an inhomogeneous IGM, which is partially ionized by these

524: sources \cite{Bromm2}.  It would be interesting to extract particular

525: realizations of the perturbed IGM around massive galaxies from numerical

526: simulations, and to apply a suitable radiative transfer code in propagating

527: the Ly$\alpha$ photons from the embedded galaxies. Mapping of the

528: properties of the associated Ly$\alpha$ halos will allow to assess their

529: detectability with future observations.

530: 

531: 

532: \begin{figure}[htbp]

533: \special{psfile=fig8.eps hoffset=-4

534: voffset=-375 hscale=60 vscale=60}

535: \vspace{3.05in}

536: \caption{{\it Loeb--Rybicki halos:}\/ Scattering of \lya line photons

537: from a galaxy embedded in the neutral intergalactic medium prior to

538: reionization. The line photons diffuse in frequency due to the Hubble

539: expansion of the surrounding medium and eventually redshift out of

540: resonance and escape to infinity.  A distant observer sees a \lya halo

541: surrounding the source, along with a characteristically asymmetric

542: line profile. The observed line should be broadened and redshifted by

543: about one thousand ${\rm km~s^{-1}}$ relative to other lines (such as

544: H$_\alpha$) emitted by the galaxy.  }

545: \label{fig8}

546: \end{figure}

547: 

548: 

549: 

550: \subsection{Magnetization of the Intergalactic Medium by Quasar Outflows}

551: 

552: Magnetic fields and cosmic rays play an important dynamical role in the

553: interstellar medium of galaxies. However, they are often ignored in

554: discussions of the intergalactic medium (IGM). {\it How significant is the

555: pressure from these non-thermal components in the IGM? Has the magnetic

556: field observed in collapsed objects, such as galaxies or clusters of

557: galaxies, originated from the IGM?}  The first question is particularly

558: relevant for the reconstruction of the mass distribution of galaxy clusters

559: from X-ray data.

560: 

561: Outflows from quasars inevitably pollute the IGM with magnetic fields. The

562: short-lived activity of a quasar leaves behind an expanding magnetized

563: bubble in the IGM. Furlanetto \& Loeb \cite{Furlanetto} modeled the

564: expansion of the remnant quasar bubbles and calculated their distribution

565: as a function of size and magnetic field strength at different redshifts.

566: They have found that generically by a redshift $z\sim 3$, about 5--$20\%$

567: of the IGM volume is filled by magnetic fields with an energy density $\ga

568: 10\%$ of the mean thermal energy density of a photo-ionized IGM at $\sim

569: 10^4$ K (see Figure~\ref{fig9}). As massive galaxies and X-ray clusters

570: condense out of the magnetized IGM, the adiabatic compression of the

571: magnetic field could result in the field strength observed in these systems

572: without a need for further dynamo amplification.  The intergalactic

573: magnetic field could also provide a non--thermal contribution to the

574: pressure of the photo-ionized gas that may account for the claimed

575: discrepancy between the simulated and observed Doppler width distributions

576: of the Ly$\alpha$ forest at $z\ga 2$ \cite{Bryan,Theuns}. However, the

577: supplied magnetic energy is unlikely to be dynamically important today

578: since the present-day IGM was heated by gravitationally-induced shocks to

579: an average mass-weighted temperature of $\sim 3\times 10^6$K \cite{Dave},

580: larger by two orders of magnitude than the magnetic energy input from

581: quasars.

582: 

583: 

584: \begin{figure}[htbp]

585: \special{psfile=fig9.ps hoffset=20 voffset=-350 hscale=50 vscale=50}

586: \vspace{3.9in}

587: \caption{{\it Upper Panel:} Volume filling fraction of magnetized bubbles

588: $F(z)$, as a function of redshift (taken from Furlanetto \& Loeb 2001).

589: {\it Lower Panel:} Ratio of magnetic energy density,

590: $\bar{u}_B/\epsilon_{-1}$, to a fiducial thermal energy density $u_{fid} =

591: 3 n(z) k T_{IGM}$ of a photoionized IGM, where $T_{IGM} = 10^4 \kel$, as a

592: function of redshift.  The solid curves assume a different minimum mass for

593: a galaxy out of which a quasar outflow may originate, $M_{h,min}$. Examples

594: include a case where this minimum mass is determined by atomic cooling

595: before reionization and by infall suppression afterward (top curve), and

596: fixed-value cases with $M_{h,min} = 10^9 \msun$ (middle curve), and

597: $M_{h,min} = 10^{10} \msun$ (bottom curve). The vertical dotted line

598: indicates the assumed redshift of reionization, $z_r=7$.}

599: \label{fig9}

600: \end{figure}

601: 

602: 

603: \subsection{Particle Acceleration in Magnetized Intergalactic Shocks}

604: 

605: Even though the present-day magnetic pressure may not be dynamically

606: important, the magnetization of the IGM by quasar outflows has important

607: consequences.  Once magnetized, the shocks produced in the IGM during the

608: formation of large-scale structure can accelerate a population of highly

609: relativistic electrons by the Fermi mechanism, similarly to the

610: collisionless shocks of supernova remnants. The accelerated electrons emit

611: synchrotron radiation as they gyrate in the embedded magnetic fields and

612: produce $\gamma$-ray radiation by inverse--Compton scattering the microwave

613: background photons. This non-thermal radiation spans some $\sim 20$ orders

614: of magnitude in photon energies from the radio to the TeV $\gamma$-ray

615: regime. The radiation is expected to delineate the strong shocks formed in

616: the cosmic web of large scale sheets and filaments in the IGM.  The

617: brightest emission originates from the strongest intergalactic shocks

618: around galaxy clusters.

619: 

620: \subsection{Non-Thermal Radio and Gamma-Ray Emission}

621: 

622: More than a third of all X-ray clusters with luminosities $\ga 10^{45}~{\rm

623: erg~s^{-1}}$ possess diffuse radio halos \cite{Giovannini}.  Based on

624: energy arguments and circumstantial evidence, these radio halos are

625: believed to be caused by synchrotron emission from shock-accelerated

626: electrons during the merger events of their host clusters

627: (\cite{Harris,Tribble,Feretti}; see \cite{Liang} for references to

628: alternative, less successful models).  These highly-relativistic electrons

629: cool primarily through inverse-Compton scattering off the microwave

630: background. Since their cooling time is much shorter than the dynamical

631: time of their host cluster, the radio emission is expected to last as long

632: as the shock persists and continues to accelerate fresh electrons to

633: relativistic energies. Intergalactic shocks also occur along the filaments

634: and sheets that channel mass into the clusters. These structures, also

635: traced by the galaxy distribution \cite{Doroshkevich}, are induced by

636: gravity and form due to converging large-scale flows in the IGM.

637: 

638: The intergalactic synchrotron emission contaminates the cosmic microwave

639: background (CMB) anisotropies -- relic from the epoch of recombination, and

640: as such needs to be considered in the design and analysis of anisotropy

641: experiments at low frequencies.  Previous estimates of synchrotron

642: contamination of the cosmic anisotropies focused on Galactic emission,

643: which occurs primarily at low Galactic latitudes and large angular scales

644: \cite{Tegmark2}.  Waxman \& Loeb \cite{Waxman_Loeb} calculated the

645: intergalactic synchrotron contribution to the fluctuations in the radio sky

646: as a function of frequency and angular scale (see Figure~\ref{fig10}). They

647: assumed that most of the emission originates from the virialization shocks

648: around X-ray clusters, and used the Press--Schechter \cite{PS} mass

649: function to describe the abundance of such clusters as a function of

650: redshift (see also subsequent work in \cite{Fujita}).  Although the

651: synchrotron background amounts to only a small fraction of the CMB

652: intensity, Waxman \& Loeb \cite{Waxman_Loeb} showed that its fluctuations

653: could dominate over the primordial CMB fluctuations at low photon

654: frequencies, $\nu \la 10$~GHz.  They found that radio emission from cluster

655: shocks contributes a fluctuation amplitude of $\sim 40 \mu$K $\times

656: (\nu/10{\rm GHz})^{-3}$ to the CMB on angular scales between 1 and

657: $0.1^\circ$, respectively.  Interestingly, current anisotropy experiments

658: are just sensitive to this level of fluctuations.

659: %\footnote{See table summary of current experiments at

660: %http://www.hep.upenn.edu/$\~$max/index.html, or at

661: %http://cfa-www.harvard.edu/$\~$mwhite/cmbexptlist.html}.  

662: Existing detections by CAT ($50\pm15\mu$K at 15~GHz on 0.2--0.5$\arcdeg$

663: scales) and OVRO ($56^{+8.5}_{-6.6}\mu$K at 20~GHz on 0.1--0.6$\arcdeg$

664: scales), as well as 95\% upper limits ($\la 40\mu$K on arcminute scales at

665: 9-15 GHz by the ATCA, RYLE and VLA detectors) are consistent with our

666: prediction.

667: %\footnote{Note that although the cluster contribution declines

668: %rapidly at high frequencies, some of these experiments, like CAT or OVRO,

669: %measured fluctuations at only one frequency and could not reject a

670: %synchrotron contribution to the measured signal.}.

671: 

672: \begin{figure}[t]

673: \special{psfile=fig10.eps hoffset=20 voffset=-300 hscale=50 vscale=50}

674: \vspace{2.9in}

675: \caption{Fractional intensity fluctuations, $(\langle

676: I_\nu(0)I_\nu(\theta)\rangle-\langle I_\nu\rangle^2)^{1/2}/ \langle I_\nu

677: \rangle$, in synchrotron (radio) and inverse-Compton ($\gamma$-ray)

678: background flux (taken from Waxman \& Loeb 2000).  The dimensionless

679: coefficient, $f_{\rm sh}$, is of order unity. The ratio of synchrotron to

680: CMB intensity is $\langle I^{\rm syn}_\nu\rangle/I^{\rm

681: CMB}_\nu=6\times10^{-6} f_{\rm sh}^{-2}(\xi_B/0.01)(\nu/{10\rm GHz})^{-3}$,

682: where the post-shock magnetic energy fraction $\xi_B$ is assumed to be of

683: order 0.01, corresponding to a magnetic field strength of 0.1$\mu$G in the

684: virialization shock of X-ray clusters.  The post-shock energy fraction

685: carried by relativistic electrons is assumed to be $\xi_e\sim 0.05$,

686: similar to the value inferred in the shocks of supernova remnants.}

687: \label{fig10}

688: \end{figure}

689: 

690: Loeb \& Waxman \cite{Loeb_Waxman} have shown that a significant fraction of

691: the diffuse $\gamma$--ray background \cite{Sreekumar} might have been

692: generated by the shocks resulting from the formation of large-scale

693: structure in the intergalactic medium. Similarly to the collisionless

694: shocks of supernova remnants \cite{Blandford}, these shocks produce a

695: population of highly-relativistic electrons with a maximum Lorentz factor

696: $\ga 10^7$ that scatter a small fraction of the microwave background

697: photons in the present-day universe up to $\gamma$-ray energies, thereby

698: providing the $\gamma$-ray background.  The predicted flux agrees with the

699: observed diffuse background over more than four decades in photon energy

700: and is not sensitive to the precise magnetic field value, provided that the

701: fraction of shock energy carried by relativistic electrons is

702: $\xi_e\sim0.05$ (a value consistent with that inferred for supernovae

703: shocks).  The same electrons that emit $\gamma$-rays by inverse-Compton

704: scattering of microwave background photons, also produce synchrotron

705: radiation in the radio band due to intergalactic magnetic fields. The

706: existence of magnetic fields with an amplitude $\ga 0.1 \mu$G, is inferred

707: in cluster halos \cite{Kim,FuscoFemiano,Rephaeli,Kaastra}, and is also

708: required for the Fermi acceleration of these electrons.  The appearance of

709: radio halos around young X-ray clusters is therefore a natural consequence

710: and an important test of this model for the extragalactic $\gamma$-ray

711: background. The combination of radio and $\gamma$-ray data can be used to

712: calibrate $\xi_e$ and determine the strength of the intergalactic magnetic

713: field.

714: 

715: The assumed acceleration of electrons by collisionless shocks is similar to

716: that observed in supernova remnants.  Recent X-ray \cite{Koyama1,Koyama2}

717: and TeV \cite{Tanimori,Muraishi} observations of the supernova remnants

718: SN1006 and SNR RX J1713.7--3946 imply that electrons are accelerated in the

719: remnant shocks up to an energy $\sim 100$~TeV, and are confined to the

720: collisionless fluid by magnetic fields.  These shocks have a velocity of

721: order $10^3~{\rm km~s^{-1}}$, similar to the velocity of the intergalactic

722: shocks we consider here.  Although the plasma density is very different in

723: the two cases, the density may be scaled out of the problem by measuring

724: time in units of $\omega_{pe}^{-1}$, where $\omega_{pe}$ is the electron

725: plasma frequency (shock characteristics may also depend on the pre-shock

726: magnetic field, which introduces a dimensionless parameter into the

727: problem, $\omega_{ce}/\omega_{pe}$, where $\omega_{ce}$ is the electron

728: cyclotron frequency. However, for both supernova and IGM shocks,

729: $\omega_{ce}/\omega_{pe}\ll1$, suggesting a similar behavior in both

730: cases).  For supernova shocks, the inferred energy density in relativistic

731: electrons constitutes $1$--$10\%$ of the post-shock energy density in these

732: remnants \cite{Ellison2}, a fraction consistent with the global ratio

733: between the mean energy density of cosmic-ray electrons and the turbulent

734: energy density in the interstellar medium of our galaxy.

735: 

736: The production of anisotropic backgrounds of radio and $\gamma$-ray

737: radiation by strong intergalactic shocks is a natural consequence of

738: structure formation in the Universe. The brightest emission originates from

739: the shocks on Mpc scales around newly formed, massive X-ray clusters.  The

740: foreground synchrotron fluctuations might be comparable to the anisotropy

741: signals detected by existing low-frequency microwave background

742: experiments, and can be easily isolated through multi-frequency

743: observations. Polarization anisotropy experiments could then constrain the

744: coherence length of the intergalactic magnetic field.

745: 

746: Different astrophysical sources may contribute to the strength of the

747: intergalactic magnetic fields; aside from radio sources

748: \cite{Daly90,Furlanetto} they include: the large scale shocks themselves

749: \cite{Kulsrud97}, the first generation of stars \cite{Rees87} and

750: supernova-driven winds from galaxies \cite{Kronberg99}.  Different

751: contributions may lead to different scalings of the magnetic field energy

752: density with IGM gas parameters.  The non-thermal $\gamma$-ray emission is

753: only weakly dependent on the magnetic field strength

754: \cite{Loeb_Waxman}. However, the non-thermal radio emission is highly

755: sensitive to this parameter and different models may lead to a different

756: dependence of the radio emission on scale and redshift, which may have

757: observable consequences (e.g. modifying the functional dependence of the

758: correlation function shown in Figure \ref{fig10}).

759: 

760: Recent observations of the giant radio galaxy NGC 315 may constitute the

761: first direct detection of collisionless large scale structure shocks,

762: accelerating electrons to high energies \cite{Ensslin01}.  Radio

763: observations of this source are most naturally explained by the existence

764: of a large scale shock produced by a flow converging towards an

765: intersection of galaxy filaments \cite{Ensslin01}.

766: 

767: Waxman \& Loeb \cite{Waxman_Loeb} predicted a fluctuation amplitude $\ga

768: 40\%$ in the $\gamma$-ray background intensity on sub-degree scale, and the

769: existence of extended, $\ga 1\arcdeg$, $\gamma$-ray halos, associated with

770: newly formed massive clusters. On scales larger than a degree the

771: fluctuation amplitude declines and is well below the anisotropy limits from

772: EGRET (see Figure 5 in \cite{Sreekumar}).  Detection of the predicted

773: signals will provide a calibration of the uncertain model parameter,

774: $\xi_e$.  The high-energy maps required to detect the predicted anisotropy

775: signal will be made between 20 MeV and 300 GeV by the GLAST instrument

776: (planned for launch in 2005; see http://glast.gsfc.nasa.gov/), which is

777: expected to be more sensitive than EGRET by an order-of-magnitude

778: \cite{Bloom96}.  The predicted $\gamma$-ray halos may constitute a

779: significant fraction of the unidentified extra-Galactic EGRET sources

780: \cite{Totani1,Totani2}.  However, since the angular extension of the

781: brightest halos is large, a more careful analysis is required in order to

782: assess the detectability of such halos by EGRET.

783: 

784: A future, dedicated, all-sky anisotropy experiment, operating at several

785: frequencies below 10 GHz, would be able to map the fluctuations in the

786: intergalactic synchrotron background.  The resulting synchrotron map could

787: then be cross-correlated with full-sky maps at hard X-ray or $\gamma$-ray

788: energies to confirm its cosmic origin. Identification of the synchrotron

789: fluctuations together with their counterpart inverse-Compton emission of

790: hard X-rays or $\gamma$-rays by the same population of shock-accelerated

791: electrons, can be used to empirically determine the strength and spatial

792: distribution of the intergalactic magnetic field.  Similarly, the

793: correlation between radio and $\gamma$-ray halos may be detectable around

794: individual X-ray clusters.  Strong radio halos could be the best indicators

795: for bright $\gamma$-ray clusters, which would provide the first obvious

796: targets for GLAST.

797: 

798: Semi-analytic models \cite{Waxman_Loeb} identify the intergalactic shocks

799: with smooth spherical accretion of gas onto clusters, while in reality they

800: result from asymmetric mergers as well as from converging flows in large

801: scale sheets and filaments. The more realistic emission from these complex

802: geometries can be best modeled through detailed hydrodynamic simulations

803: \cite{Miniati1,Miniati3,Keshet}. Mergers of comparable mass clusters, for

804: example, would tend to produce only mild shocks due to the prior heating of

805: the shocked gas, leading to reduced non-thermal emission due to the steep

806: power-law slope of the accelerated electrons.

807: 

808: In analogy with the collisionless shocks of supernova remnants, the

809: intergalactic shocks are also expected to accelerate a hadronic cosmic-ray

810: component that would acquire a fraction ($\sim 10$--$50\%$) of the

811: post-shock energy more substantial than that carried by relativistic

812: electrons ($\sim 1$--$10\%$, see \cite{Ellison2}). Collisions of the

813: hadronic cosmic-rays with IGM protons produce pions, $\pi^0$, which decay

814: into $\gamma$-ray photons; however for typical parameters this radiation

815: component is expected to be sub-dominant relative to the inverse-Compton

816: production of $\gamma$-rays by the relativistic electrons

817: \cite{Loeb_Waxman,Waxman_Loeb,Miniati2}.  The buoyancy of the relativistic

818: cosmic--ray fluid may lead to instabilities that would eventually separate

819: it from the IGM gas and allow it to expand into the large-scale voids that

820: fill most of the volume in between the sheets and filaments of the IGM. For

821: a relativistic fluid, the adiabatic decrease of the pressure with

822: increasing volume ($p\propto V^{-4/3}$) is milder than for the IGM gas

823: ($p\propto V^{-5/3}$). It is possible that the intergalactic voids are

824: currently being dominated by cosmic--ray and magnetic field pressure which

825: suppresses the formation of galaxies there.

826: 

827: 

828: \bigskip

829: \noindent{\bf Acknowledgements}

830: 

831: \noindent

832: I thank all my collaborators on the topics described in this brief review:

833: Rennan Barkana, Benedetta Ciardi, Steve Furlanetto, Peter Garnavitch, Scott

834: Gaudi, Jonathan Granot, Zoltan Haiman, Shude Mao, Rosalba Perna, George

835: Rybicki, Kris Stanek, and Eli Waxman. This work was supported in part by

836: NASA grants NAG 5-7039, 5-7768, and by NSF grants AST-9900877, AST-0071019.

837: 

838: \vfil\eject

839: \begin{thebibliography}{8.}

840: \addcontentsline{toc}{section}{References}

841: 

842: \bibitem{Abel} T. Abel, G. Bryan, M. L. Norman: Ap.~J., \textbf{540}, 39

843: (2000)

844: 

845: \bibitem{Anderson} A. I. Anderson, et al.: A. \& A., \textbf{364}, L54

846: (2000)

847: 

848: \bibitem{Barkana} R. Barkana, A. Loeb: Physics Reports, \textbf{349}, 125

849: (2001)

850: 

851: \bibitem{Becker} R. H. Becker et al: Astr.~J., \textbf{submitted};

852: astro-ph/0108097 (2001)

853: 

854: \bibitem{Berger} E. Berger, et al.: Ap.~J., \textbf{545}, 56 (2000)

855: 

856: \bibitem{Blaes} O. M. Blaes, R. L. Webster: Ap.~J., \textbf{391}, L63

857: (1992)

858: 

859: \bibitem{Blandford} R. Blandford, D. Eichler: Physics Reports,

860: \textbf{154}, 1 (1987)

861: 

862: \bibitem{Bloom96} E. D. Bloom: Space Sci. Rev., \textbf{75}, 109 (1996)

863: 

864: \bibitem{Bromm} V. Bromm, P. S. Coppi, R. B. Larson: \textbf{527}, 5

865: (1999)

866: 

867: \bibitem{Bromm2} V. Bromm, A. Loeb, G. Rybicki: in preparation (2001)

868: 

869: \bibitem{Bryan} G.L. Bryan, M. Machacek, P. Anninos, M.L.  Norman: Ap.~J.,

870: \textbf{517}, 13 (1999)

871: 

872: \bibitem{Ciardi} B. Ciardi, A. Loeb: Ap.~J., \textbf{540}, 687 (2000)

873: 

874: \bibitem{Dave} R. Dave, et al.: Ap.~J., \textbf{552}, 473 (2001)

875: 

876: \bibitem{Djorgovski} S. G. Djorgovski, S. M.  Castro, D. Stern, A. Mahabal:

877: Ap.~J.~L., \textbf{submitted}; astro-ph/0108069 (2001)

878: 

879: \bibitem{Daly90} R. A.  Daly, A. Loeb: Ap.~J. \textbf{364}, 451 (1990)

880: 

881: \bibitem{Doroshkevich} A. G. Doroshkevich et al.: M.N.R.A.S., \textbf{283},

882: 1281 (1996)

883: 

884: \bibitem{Ellison} S. L. Ellison, A. Songaila, J. Schaye, M. Pettini:

885: A. J., \textbf{120}, 1175 (2000)

886: 

887: \bibitem{Ellison2} D. C. Ellison, P. Slane, B. M. Gaensler: Ap.~J.,

888: \textbf{submitted}; astro-ph/0106257 (2001)

889: 

890: \bibitem{Ensslin01} T. A.  Ensslin et al.: Ap.~J.,

891: \textbf{549}, L39 (2001)

892: 

893: \bibitem{Fan} X. Fan, et al.: A. J., \textbf{120}, 1167 (2000)

894: 

895: \bibitem{Feretti} L. Feretti: to appear in Proc. of IAU Symp. \# 199 on

896: ``The Universe at Low Radio Frequencies'', Pune, India; astro-ph/0006379

897: (2000)

898: 

899: \bibitem{Fiore} F. Fiore, F. Nicastro, S. Savaglio, L. Stella, M.

900: Vietri: Ap.~J., \textbf{544}, L7 (2000)

901: 

902: \bibitem{Frail} D. A. Frail, et al.: Nature, \textbf{submitted} (2001);

903: astro-ph/0102282

904: 

905: \bibitem{Freedman} D. L. Freedman, E. Waxman: Ap.~J., \textbf{547}, 922

906: (2001)

907: 

908: \bibitem{Fruchter} A. Fruchter: Ap.~J., \textbf{512}, L1 (1999)

909: 

910: \bibitem{Fujita} Y. Fujita, C. L. Sarazin, Ap.~J., \textbf{in press};

911: astro-ph/0108369 (2001)

912: 

913: \bibitem{Fukugita} M. Fukugita, C. J. Hogan, P. J. E. Peebles: Nature,

914: \textbf{366}, 309 (1998)

915: 

916: \bibitem{Furlanetto} S. R. Furlanetto, A. Loeb: Ap.~J., \textbf{556}, 619

917: (2001)

918: 

919: \bibitem{Garnavich} P. M. Garnavich, A. Loeb, K. Z. Stanek: Ap.~J.,

920: \textbf{544}, L11 (2000)

921: 

922: \bibitem{FuscoFemiano} R. Fusco-Femiano et al.: Ap.~J., \textbf{513},

923: L21 (1999)

924: 

925: \bibitem{Gaudi_Loeb} B. S. Gaudi, A. Loeb: Ap.~J., \textbf{558} (2001)

926: 

927: \bibitem{GGL} B. S. Gaudi, J. Granot, A. Loeb: Ap.~J., \textbf{in press}

928: (2001)

929: 

930: \bibitem{Giovannini} G. Giovannini, M. Tordi, L. Feretti: New Astronomy,

931: \textbf{4}, 141 (1999)

932: 

933: \bibitem{Goodman} J. Goodman: New Astronomy, \textbf{2}, 449 (1997)

934: 

935: \bibitem{Granot_Loeb} J. Granot, A. Loeb: Ap.~J., \textbf{551}, L63 (2001)

936: 

937: \bibitem{Granot} J. Granot, T. Piran, R. Sari: Ap.~J., \textbf{513},

938: 679 (1999a); \textbf{527}, 236 (1999b)

939: 

940: \bibitem{Gruzinov} A. Gruzinov, E. Waxman: Ap.~J., \textbf{511}, 852 (1999)

941: 

942: \bibitem{Gunn} J. E. Gunn, B. A. Peterson: Ap.~J., \textbf{142}, 1633,

943: (1965)

944: 

945: \bibitem{Haiman} Z. Haiman, A. Loeb: Ap.~J., \textbf{483}, 21 (1997);

946: \textbf{503}, 505 (1998); \textbf{552}, 459 (2001)

947: 

948: \bibitem{Hamann} F. Hamann, G. Ferland: A. R. A. \& A. \textbf{37}, 487

949: (1999)

950: 

951: \bibitem{Harris} D. E. Harris, V. K. Kapahi, R. D. Ekers:

952: Astr. \&Ap. Suppl., \textbf{39}, 215 (1995)

953: 

954: \bibitem{Hartwick} F. D. A. Hartwick, D. Schade: A. R. A. \& A.,

955: \textbf{28}, 437 (1990)

956: 

957: \bibitem{Hellsten} U. Hellsten, N. Y. Gnedin, J. Miralda-Escude: ApJ,

958: \textbf{509}, 56 (1998)

959: 

960: \bibitem{Ioka} K. Ioka, K., T. Nakamura: Ap.~J., \textbf{in press};

961: astro-ph/0102028 (2001)

962: 

963: \bibitem{Kaastra} J. S. Kaastra, et al.: Ap.~J., \textbf{519}, L119 (1999)

964: 

965: \bibitem{Keshet} U. Keshet, E. Waxman, V. Springel, A. Loeb, L. Hernquist:

966: Ap.~J., in preparation (2001)

967: 

968: \bibitem{Kim} K.-T.  Kim, P. P. Kronberg, G. Giovannini, T.  Venturi:

969: Nature, \textbf{341}, 720 (1989)

970: 

971: \bibitem{Koopmans} L. V. E. Koopmans, J. Wambsganss: M.N.R.A.S., \textbf{in

972: press}; astro-ph/0011029 (2001)

973: 

974: \bibitem{Koyama1} K.  Koyama et al.: Nature, \textbf{378}, 255 (1995)

975: 

976: \bibitem{Koyama2} K.  Koyama et al.: PSAJ, \textbf{49}, L7 (1997)

977: 

978: \bibitem{Kronberg99} P. P. Kronberg, H. Lesch, U. Hopp: Ap.~J.,

979: \textbf{511}, 56 (1999)

980: 

981: \bibitem{Kulsrud97} R. M.  Kulsrud, R. Cen, J. P. Ostriker, D. Ryu: Ap.~J.,

982: \textbf{454}, 60 (1997)

983: 

984: \bibitem{Mao} S. Mao, A. Loeb: Ap.~J., \textbf{547}, L97 (2001)

985: 

986: \bibitem{Lamb} D.\ Q.\ Lamb, D.\ E.\ Reichart:

987: Ap.~J. \textbf{536}, 1 (2000)

988: 

989: \bibitem{Liang} H. Liang, R. W. Hunstead, M. Birkinshaw, P.  Andreani:

990: Ap.~J., \textbf{544}, 686 (2000)

991: 

992: \bibitem{Loeb_Barkana} A. Loeb, R. Barkana: ARA\&A, \textbf{in press}

993: (2001)

994: 

995: \bibitem{Loeb_Perna} A. Loeb, R. Perna: Ap.~J., \textbf{495}, 597 (1998)

996: 

997: \bibitem{Loeb_Rybicki} A. Loeb, \& G. Rybicki: Ap.~J., \textbf{524}, 527

998: (1999)

999: 

1000: \bibitem{Loeb_Waxman} A.  Loeb, E. Waxman: Nature, \textbf{405}, 156 (2000)

1001: 

1002: \bibitem{Medvedev} M. V. Medvedev, A. Loeb: Ap.~J. \textbf{526}, 697 (1999)

1003: 

1004: \bibitem{Miniati1} F. Miniati, D. Ryu, H. Kang, T. W. Jones, R. Cen,

1005: J. P. Ostriker: Ap.~J., \textbf{542}, 608 (2000)

1006: 

1007: \bibitem{Miniati2} F. Miniati, D. Ryu, H. Kang, T. W.  Jones: Ap.~J.,

1008: \textbf{559}, 1 (2001)

1009: 

1010: \bibitem{Miniati3} F. Miniati, T. W.  Jones, H. Kang, D. Ryu: Ap.~J.,

1011: \textbf{in press}; astro-ph/0108305 (2001)

1012: 

1013: \bibitem{Miralda_Escude} J. Miralda-Escud\'e: Ap.~J., \textbf{501}, 15

1014: (1998)

1015: 

1016: \bibitem{Muraishi} H. Muraishi et al.: Astr. \& Ap., \textbf{354},

1017: L57 (2000)

1018: 

1019: \bibitem{Nemiroff} R. Nemiroff: Ap.\&~SS, \textbf{259}, 309 (1998)

1020: 

1021: \bibitem{Nemiroff2} R. Nemiroff, J. P. Norris, J. T. Bonnell, 

1022: G. F. Marani: Ap.~J., \textbf{494}, L173 (1998)

1023: 

1024: \bibitem{Panaitescu} A. Panaitescu, P. M\'esz\'aros: Ap.~J., \textbf{493},

1025: L31 (1998)

1026: 

1027: \bibitem{Perna} R. Perna, A. Loeb: Ap.~J., \textbf{501}, 467 (1998)

1028: 

1029: \bibitem{Perna2} R. Perna, \& A. Loeb: Ap.~J., \textbf{503}, L135 (1998)

1030: 

1031: \bibitem{Press} W. H. Press, J. E. Gunn: Ap.~J., \textbf{185}, 397 (1973)

1032: 

1033: \bibitem{PS} W. H. Press, P. Schechter: Ap.~J., \textbf{187}, 425 (1974)

1034: 

1035: \bibitem{Rees87} M. J.  Rees: QJRAS, \textbf{28}, 197 (1987)

1036: 

1037: \bibitem{Rephaeli} Y. Rephaeli, D. Gruber, P.  Blanco: Ap.~J.,

1038: \textbf{511}, L21 (1999)

1039: 

1040: \bibitem{Rybicki} G. Rybicki, \& A. Loeb: Ap.~J., \textbf{520}, L79 

1041: (1999)

1042: 

1043: \bibitem{Sari} R. Sari: Ap.~J., \textbf{494}, L49 (1998)

1044: 

1045: \bibitem{Schneider} P. Schneider, J. Ehlers, E. E. Falco:

1046: \emph{Gravitational Lenses} (Springer, Heidelberg 1992)

1047: 

1048: \bibitem{Sreekumar}

1049: P. Sreekumar et al.: Ap.~J., \textbf{494}, 523 (1998)

1050: 

1051: \bibitem{Stern} D. Stern, H. Spinrad: P. A. S. P., \textbf{111}, 1475

1052: (1999)

1053: 

1054: \bibitem{Tanimori} T. Tanimori et al.: Ap.~J., \textbf{497}, L25

1055: (1998)

1056: 

1057: \bibitem{Tegmark} M. Tegmark, J. Silk, A. Evrard: Ap.~J., \textbf{417}, 54

1058: (1993)

1059: 

1060: \bibitem{Tegmark2} M. Tegmark, D.J. Eisenstein, W. Hu, A. de

1061: Oliveira-Costa: Ap.~J., \textbf{530}, 133 (2000)

1062: 

1063: \bibitem{Theuns}

1064: T. Theuns, A. Leonard, G. Efstathiou, F.R. Pearce, P.R. Thomas:

1065: M.N.R.A.S., \textbf{301}, 478 (1998)

1066: 

1067: \bibitem{Totani1} T. Totani, T. Kitayama: Ap.~J., \textbf{545}, 572 (2000)

1068: 

1069: \bibitem{Totani2} W. Kawasaki, T. Totani: Ap.~J., \textbf{submitted};

1070: astro-ph/0108309 (2001)

1071: 

1072: \bibitem{Tribble} P. Tribble:  M.N.R.A.S., 263, 31 (1993)

1073: 

1074: \bibitem{Wang} X. Wang, M. Tegmark, M. Zaldarriaga: Phys.~Rev.~D,

1075: \textbf{submitted}; astro-ph/0105091 (2001)

1076:  

1077: \bibitem{Waxman} E. Waxman: Ap.~J., \textbf{491}, L19 (1997)

1078: 

1079: \bibitem{WKF} E. Waxman, S. R. Kulkrani, D. A. Frail: Ap.~J., \textbf{497},

1080: 288 (1998)

1081: 

1082: \bibitem{Waxman_Loeb} E. Waxman, A. Loeb: Ap.~J., \textbf{545}, L11 (2000)

1083: 

1084: \bibitem{Zhang} Z. Zhang, P. Meszaros: Ap.~J., \textbf{552}, L35 (2001)

1085: 

1086: \end{thebibliography}

1087: 

1088: \end{document}

1089: