hep-ph0102106/z3.tex
1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 
2: %% Revtex document %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 
3: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 
4: 
5: \documentstyle[aps,prl,epsfig,floats]{revtex} 
6: \bibliographystyle{apsrev}
7: 
8: \flushbottom
9: \renewcommand{\textfraction}{0}
10: \renewcommand{\topfraction}{1}
11: \renewcommand{\bottomfraction}{1}
12: 
13: \draft
14:  
15: \begin{document} 
16: 
17: \twocolumn[\hsize\textwidth\columnwidth\hsize\csname
18: @twocolumnfalse\endcsname 
19: 
20: 
21: \title{Neutrinos as a signature of ultrahigh energy photons at high red
22: shift} 
23:  
24: \author{Marieke Postma} 
25: \address{Department of Physics and Astronomy, UCLA, Los Angeles, 
26: CA 90095-1547
27: }           
28:  
29: \date{\today}
30: 
31: \maketitle
32:  
33: \begin{abstract}
34: Sources of ultrahigh energy photons operating at high red shift
35: produce a diffuse background of neutrinos.  At high red shift, when
36: the cosmic microwave background radiation has a higher temperature, an
37: electromagnetic cascade originated by an energetic photon can generate
38: neutrinos via muon and pion production and decay.  We have calculated
39: numerically the neutrino spectrum produced by various photon sources.
40: A distinctive feature of the produced flux is a ``bump'' in the
41: spectrum at neutrino energies $E \sim 10^{17}$eV.  The produced flux
42: is largest for $m=3$ sources ({\it e.g} necklaces), with $E^2 J(E)
43: \sim 10\, {\rm eV} {\rm cm}^{-2} {\rm s}^{-1} {\rm sr}^{-1}$ at these
44: energies.  Detection of such neutrinos can help understand the origin
45: of ultrahigh energy cosmic rays.
46: \end{abstract} 
47:  
48: \pacs{PACS numbers: 98.70.Sa, 95.85.Ry, 98.70.Vc 
49: \hspace{1.0cm}  UCLA/01/TEP/2
50: }
51: 
52:  
53: \vskip2.0pc]
54: 
55: \renewcommand{\thefootnote}{\arabic{footnote}}
56: \setcounter{footnote}{0}
57: 
58:  
59: \section{Introduction}
60: 
61: In the near future several experiments will become sensitive to small
62: fluxes of very high energy neutrinos.  One may hope to observe
63: neutrinos from active galactic nuclei (AGN), gamma ray bursts, other
64: astrophysical sources, and, possibly, from yet undiscovered objects
65: that have emerged from particle theory, {\em e.g.}, topological
66: defects (TD).  To learn new physics from the future observations, it
67: is important to identify signatures of certain sources that may
68: contribute to diffuse neutrino flux.  In this paper we detail the
69: description of one such signature~\cite{kp}, namely the imprint of
70: ultrahigh energy photons from high red shift on the neutrino spectrum.
71: At high red shift, when the cosmic microwave background radiation
72: (CMBR) has a higher temperature, an electromagnetic cascade originated
73: by an ultrahigh energy photon can generate neutrinos via muon and pion
74: production and decay.  The present-day energies of such neutrinos are
75: about $10^{17}$~eV.  Detection of a ``bump'' in the neutrino spectrum
76: at $10^{17}$~eV can indicate the presence of ultrahigh energy photon
77: sources at high red shift.
78: 
79: The question of ultrahigh energy photon sources was brought in sharp
80: focus by recent observations of cosmic rays~\cite{data} with energies
81: beyond the Greisen-Zatsepin-Kuzmin (GZK) cutoff~\cite{gzk}.  The
82: origin of such cosmic rays remains an outstanding
83: puzzle~\cite{reviews}. Most of the proposed explanations can be
84: categorized in two main classes.  In the {\em bottom-up/acceleration}
85: scenarios, charged particles are accelerated to ultrahigh energies in
86: giant astrophysical ``accelerators'', such as active galactic nuclei
87: and radio galaxies~\cite{conv,mpr}.  In the {\em top-down/decay}
88: scenario on the other hand, massive objects such as topological
89: defects~\cite{TD_review,TD,TD_ber,TD_nu} and superheavy relic
90: particles~\cite{particles,gk2,kt} decay, emitting mainly ultrahigh
91: energy photons.  Topological defects and relic particles can exist at
92: much higher red shift than the astrophysical candidate sources, which
93: are formed only after the onset of galaxy formation.
94: 
95: To understand the origin of the ultrahigh energy cosmic rays (UHECR),
96: it is crucial to distinguish between these two different scenarios.
97: In~\cite{kp} we showed that sources of ultrahigh energy photons
98: operating at high red shift produce a diffuse background of neutrinos
99: with energies $E_\nu \sim 10^{17}$eV.  Our rough estimates indicated
100: that for some photon sources, {\it e.g.} necklaces, this neutrino flux
101: is large enough to be detectable in the near future.  It is therefore
102: useful to get a better estimate of the produced neutrino flux.  To
103: this end we have done a numerical calculation.  In the present paper
104: we will present the results.
105: 
106: Generation of ultrahigh energy neutrinos has been
107: studied~\cite{TD,TD_ber,TD_nu,conv,mpr,hs,stecker,ysl,pj,ps,whitepaper}
108: for various sources at small red shift, for which muon pair production
109: can be neglected.  However, a substantial flux of neutrinos could be
110: produced at earlier times, when the propagation of photons was
111: different from that in the present universe because the intergalactic
112: magnetic field was weaker, the density of radio background was lower,
113: and the cosmic microwave background density and temperature were
114: higher.  At red shift $z$ the temperature of the cosmic microwave
115: background radiation increases by a factor $(1+z)$.  Because of this,
116: at high red shift ultrahigh energy photons and electrons can produce
117: muons through interactions with the CMB photons. Muons decay into
118: neutrinos.  This is in sharp contrast with the $z \lesssim 1$ case,
119: where the photons do not produce neutrinos as they lose energy mainly
120: by scattering off the radio background and in the subsequent
121: electromagnetic cascade~\cite{berezinsky,reviews}.
122: 
123: In section~\ref{production} we will describe in detail this neutrino
124: production mechanism.  We will give an estimate for the produced flux
125: of neutrinos in section~\ref{estimates}; the numerical results are
126: presented in section~\ref{numerics}.  Our discussion applies to any
127: source of photons active at high red shift.  The latter requirement
128: excludes most astrophysical sources~\cite{conv}.  Topological defects
129: and decaying relic particles on the other hand, could operate at much
130: higher red shifts.  These sources are expected to produce photons with
131: energies as high as $10^{20}$eV.  Their presence need not be connected
132: to the GZK puzzle.  Superconducting strings and other rapidly evolving
133: topological defects, for example, are ruled out as the origin of
134: observed ultrahigh energy cosmic rays, because they would produce too
135: large a flux of secondary, low-energy photons.  However, despite this
136: constraint on their present density, such sources could have existed
137: in the early universe and could produce a detectable flux of
138: neutrinos.
139: 
140: 
141: \section{Neutrino production}
142: \label{production}
143: 
144: The propagation of energetic photons and electrons is governed by
145: their interactions with the cosmic background radiation.  At small red
146: shift they lose energy mainly through interactions with the radio
147: background. The main interactions involved in the electromagnetic
148: cascade are pair production $\gamma \gamma \to e^+e^-$ and inverse
149: Compton scattering $e \gamma \to e \gamma$.  The cross section for
150: pair production is peaked near the threshold, where $E_\gamma \epsilon
151: \simeq m_e^2$.  Therefore, for ultrahigh energy (UHE) photons the most
152: effective targets are background photons with energy $\epsilon \approx
153: m_e^2/E_\gamma \lesssim 10^{-6} {\rm eV}$, i.e., radio photons. The
154: radio background is generated by normal and radio galaxies.  Its
155: present density~\cite{biermann} is higher than that of CMB photons in
156: the same energy range.  As a result, the mean free path for energetic
157: photons and electrons is determined by the radio background, and at
158: small red shift the main source of energy loss for UHE photons is
159: electromagnetic cascade generated by interactions with radio photons.
160: At red shift $z$, however, the density of CMB photons is higher by a
161: factor $(1+z)^3$, while the density of radio background is either
162: constant or, more likely, lower.  Some models of cosmological
163: evolution of radio sources~\cite{condon} predict a sharp drop in the
164: density of radio background at red shift $z \gtrsim 2$.  More recent
165: observations~\cite{recent} indicate that the decrease of radio
166: background at $z>2$ is slow.  In any case, at high red shift one
167: expects the CMB to set the mean free path and thus become the dominant
168: source of energy loss for photons, because of the $(1+z)^3$ increase
169: in the density.  Let $z_{_{\rm R}}$ be the red shift at which
170: scattering off radio background can be neglected with respect to
171: scattering of CMBR.  Based on the analyses of
172: Refs.~\cite{condon,recent}, we take $z_{_{\rm R}} \sim 5$.
173: 
174: Another source of energy loss is the synchrotron radiation emitted by
175: the cascade electrons in the intergalactic magnetic field (IGMF).
176: This is an important effect at small red shift, but it becomes less
177: significant at earlier times when the IGMF is weak.  For $z>z_{_{\rm
178: M}} \sim 5$ synchrotron losses are small compared to energy losses
179: resulting from interactions with the CMB radiation. We will use the
180: value $z_{\rm min}=\max (z_{_{\rm R}}, z_{_{\rm M}}) \approx 5$ in what
181: follows.  As discussed in section~\ref{numerics}, a higher value of
182: $z_{\rm min}$, even as high as 10, would not make a big difference in
183: the flux of the signature neutrinos.
184: 
185: Not only the density of the CMB increases with red shift, so does its
186: temperature: $T_{_{CMB}}(z) = 2.7 (1+z){\rm K}$.  The center of mass
187: energy in scatterings of UHE electrons and photons with CMB photons
188: increases, and new, neutrino producing interactions become possible.
189: At red shift $z>z_{\rm min}$, where radio background and the
190: intergalactic magnetic field can be neglected, photons and electrons
191: scattering off the CMB can produce muons through the processes $\gamma
192: \gamma_{_{CMB}} \rightarrow \mu^+ \mu^-$ and $e \gamma_{_{CMB}}
193: \rightarrow e \mu^+ \mu^- $. Muons decay into neutrinos: $\mu
194: \rightarrow e \nu_e \nu_{\mu}$.  The threshold for these muon
195: producing interactions is $\sqrt{s}> 2 m_\mu=0.21$GeV, or
196: %
197: \begin{equation} 
198: E_{\gamma,e} > E_{\rm th}(z)=\frac{ 10^{20}{\rm eV}}{1+z} 
199: \label{threshold} 
200: \end{equation} 
201: %
202: 
203: But will muons, and thus neutrinos, indeed be produced? To answer this
204: question one has to look closely at the propagation of UHE photons at
205: $z \gtrsim 5$.  Photons will scatter off the CMBR; for photon energies
206: above the threshold for muon pair production, the reactions $\gamma
207: \gamma_{_{CMB}}\rightarrow e^+ e^-$, $\; \gamma
208: \gamma_{_{CMB}}\rightarrow e^+ e^- e^+ e^-$ and $\gamma
209: \gamma_{_{CMB}}\rightarrow\mu^+ \mu^-$ are possible.  For $\sqrt{s}>
210: 2m_{\pi^{\pm}} = 0.28$GeV charged pion production may also occur.
211: Accelerator experiments show that the pion cross section is small
212: compared to the cross sections for muon production at the energies of
213: interest~\cite{pi}. We will, therefore, neglect it.
214: 
215: The cross section for electron pair production (PP) decreases with
216: increasing photon energy
217: %
218: \begin{equation}
219:   \sigma_{\rm PP} = \frac{8 \pi \alpha^2}{s} \ln \frac{s}{m_e^2} 
220:   \qquad (s \gg m_e^2), 
221: \label{sig_PP}
222: \end{equation}
223: %
224: while the cross section for double pair production (DPP), a higher
225: order QED process, quickly approaches its asymptotic value~\cite{dpp}
226: %
227: \begin{equation}
228: \sigma_{_{\rm DPP}} \simeq  \frac{172 \alpha^4}{36 \pi m_e^2} 
229: 	\simeq 6.45 \mu{\rm barn} \qquad (s \gg m_e^2). 
230: \label{dpp} 
231: \end{equation}
232: %
233: PP is the dominant reaction for photon energies $E_\gamma \lesssim 5
234: \times 10^{20} {\rm eV} /(1+z)$.  Since the energies of the two
235: interacting photons are vastly different, either the electron or the
236: positron from PP has energy close to that of the initial photon.  At
237: higher photon energies, DPP becomes more important.  One or more
238: energetic electrons are produced in this reaction.  Muon production is
239: suppressed at all energies, its cross section being smaller than that
240: for electron pair production by a factor $2$ at threshold to a factor
241: $10$ at higher energies. Thus after an initial $\gamma
242: \gamma_{_{CMB}}$ reaction, there is a small chance muons are produced,
243: but most likely one ends up with one or more UHE electrons.
244: 
245: These electrons continue to scatter off the CMBR.  At lower energies,
246: inverse Compton scattering, $e \gamma_{_{CMB}}\rightarrow e \gamma$,
247: converts high-energy electrons into high-energy
248: photons~\cite{reviews}.  However, at energies above the muon
249: threshold, higher order processes, such as triplet production (TPP) $e
250: \gamma_{_{CMB}}\rightarrow e\, e^+ e^-$ and electron muon-pair
251: production (MPP) $e \gamma_{_{CMB}}\rightarrow e\, \mu^+ \mu^-$,
252: dominate.  For higher energies charged pion production may also occur,
253: but we expect it to be suppressed by the same hadronic physics that
254: suppressed pion production in the photon-photon reaction, and we will
255: neglect it.  For center of mass energies $s \gg m_e^2$, the
256: inelasticity $\eta $ for TPP is very small~\cite{reviews,tpp,tpp2}:
257: %
258: \begin{equation}
259: \eta \simeq 1.768 (s/m_e^2)^{-3/4} < 10^{-3}.
260: \label{eta_tpp}
261: \end{equation}
262: %
263: Hence, the energy attenuation length $\lambda_{\rm eff}$ is much
264: larger than the TPP interaction length: $\lambda_{_{\rm TPP}} \simeq
265: \eta \lambda_{\rm eff}$.  One of the electrons produced through TPP
266: carries almost all ($1-\eta$) of the incoming electron's energy.  It
267: can interact once again with the CMBR. As a result, the leading
268: electron can scatter many times before losing a considerable amount of
269: energy.  Each time it scatters with a CMB photon there is another
270: chance to produce a muon pair, until the electron energy decreases
271: below threshold for muon production.  To determine whether muons are
272: produced, one must compare the {\em energy attenuation} length for
273: triplet production with the {\em interaction length} for muon-pair
274: production.  The interaction length is given by $ \lambda^{-1} \simeq
275: \left< n_{_{CMB}} \right> v \sigma$, and thus the ratio is
276: $R=\lambda_{\rm eff}/\lambda_{_{\rm MPP}} \simeq \sigma_{_{\rm
277: MPP}}/(\eta \sigma_{_{\rm TPP}})$. For $s \gg m_e^2$ the cross section
278: for TPP is~\cite{tpp,tpp2}
279: %
280: \begin{equation}  
281: \sigma_{_{TPP}} \simeq \frac{\alpha^3}{m_e^2} \left( \frac{28}{9} {\rm
282: ln}\frac{s}{m_e^2} -\frac{218}{27} \right).
283: \label{sig_TPP}
284: \end{equation}
285: %  
286: The MPP cross section in the energy range just above the threshold $5
287: m_\mu^2<s<20 m_\mu^2$ is of the order of $0.1-1{\rm mb}$, and the
288: ratio in this range is $R \sim 100$.
289:  
290: Since $\lambda_{\rm eff} \gg \lambda_{_{\rm MPP}}$, in the absence of
291: dense radio background and intergalactic magnetic fields, all
292: electrons with $E \gtrsim E_{\rm th}$ pair-produce muons before their
293: energy is reduced by the cascade. For muon production close to the
294: threshold, each muon carries on average $1/4$ of the incoming
295: electron's energy~\cite{tpp2}. Muons decay before they can interact
296: with the photon background.  Each energetic muon produces two
297: neutrinos and an electron.  The electron produced alongside the muon
298: pair gets half or more of the incoming electron's energy; it can
299: interact again with the CMBR to produce muons. This process can repeat
300: until the energy of the regenerated electron decreases below the
301: threshold for muon pair production.
302: 
303: 
304: 
305: \section{Flux estimates}
306: \label{estimates}
307: 
308: Let us assume that the time dependence of the rate of photon emission
309: can be parametrized as
310: %
311: \begin{equation}
312: \dot{n}_{\gamma}(E_\gamma, t) = \dot{n}_{\gamma,0}(E) \left(
313:  \frac{t}{t_0} \right)^{-m}.
314: \label{z_dep}
315: \end{equation}
316: %
317: The subscript $0$ denotes the present-day value of the parameter.  The
318: value of $m$ depends on the source under consideration.  For long
319: lived relic particles and topological defects, the rate of photon
320: emission is indeed (approximately) of the above form, with $m=0$ for
321: decaying relic particles, $m=3$ for strings, monopolonium and
322: necklaces, and $m \ge 4$ for superconducting strings.  Monopolonium is
323: a bound monopole anti-monopole pair.  Necklaces are networks of
324: strings and monopoles, with two monopoles attached to each string: the
325: monopoles are the beads in a necklace of strings.
326: 
327: All photons produced at red shifts $z_{\rm min} \leq z \leq z_{\rm
328: max}$ contribute to the present neutrino flux.  Here $z_{\rm min} \sim
329: 5$ is the minimum red shift for which radio background and magnetic
330: fields are negligible, and $z_{\rm max}$ is the maximum red shift for
331: which the universe is transparent to UHE neutrinos. The value of
332: $z_{\rm max}$ is determined by the neutrino interactions with the
333: relic neutrino background.  The absorption red shift for neutrinos
334: with energy $\sim 10^{17}$eV is $z_{\rm max}\sim 3 \times
335: 10^3$~\cite{ggs}.  The neutrino flux is:
336: %  
337: \begin{eqnarray} \label{nu.z} 
338: \phi_\nu & = & \xi \int^{z_{\rm max}}_{z_{\rm min}} {\rm d}t \,
339: \dot{n}_{\gamma}(z) \ (1+z)^{-4} \\ & = & \xi \frac{3}{-2a} t_0
340: \dot{n}_{\gamma,0}(E>E_{\rm th}) \, \, [(1+z_{\rm min})^a - (1+z_{\rm
341: max})^a], \nonumber
342: \end{eqnarray}  
343: %
344: where $a=(3m-11)/2$, and $\xi$ is the number of neutrinos produced per
345: UHE photon.  An UHE photon produces an UHE electron.  If energetic
346: enough, this electron produces a muon pair, and four neutrinos are
347: generated.  However, for initial electron energies just above
348: threshold, it may often occur that repeated TPP reactions decreases
349: the energy below threshold before any muons are produced.  Thus for
350: photons with energies close to threshold for muon production $\xi <
351: 4$.  On the other hand, for higher photon energies one can have $\xi >
352: 4$.  This is because higher energetic photons are more likely to
353: produce a double pair of electrons (DPP), {\it i.e.}, produce more
354: than one UHE electron.  Furthermore, the electron produced alongside
355: the muon pair in MPP may be energetic enough for a second round of
356: muon pair-production.  Our numerical calculation (see
357: section~\ref{numerics}) gives $\xi = 0.6$ for $E_\gamma (1+z) = 1
358: \times 10^{20}{\rm eV}$, $\;\xi = 4.0$ for $E_\gamma (1+z) = 2 \times
359: 10^{20}{\rm eV}$, and $\xi = 7.7$ for $E_\gamma (1+z) = 3 \times
360: 10^{20}{\rm eV}$.  We will take $\xi \sim 4$ in our estimate.
361: 
362: Rapidly evolving sources with $m \ge 4$, such as superconducting
363: strings, are ruled out as the origin of the observed ultrahigh energy
364: cosmic rays.  The strongest constraint on the present density of such
365: sources comes from measurements of the diffuse $\gamma$-background
366: below $100 \, {\rm GeV}$~\cite{reviews,TD_ber,egret}.  UHE photons
367: lose energy in the electromagnetic cascade involving PP and ICS.  When
368: the photon energy drops below PP threshold, the photon attenuation
369: length becomes small, and for $z \lesssim 10^3$ the universe is
370: transparent to these low energetic photons.  The flux of the cascade
371: photons in the energy range $10 \, {\rm MeV} < E <100 \, {\rm GeV}$
372: must be lower than the flux measured by the Energetic Gamma Ray
373: Experiment Telescope (EGRET): $\omega_{\rm cas} \le 2.6 \times 10^{-6}
374: {\rm eV} {\rm cm}^3$. Although the present density of $m \ge 4$
375: sources is constrained, it is conceivable that in the early universe
376: they may have existed in large enough number to produce an appreciable
377: neutrino flux.
378: 
379: For $m < 11/3$, $a < 0$, and, according to~eq.~(\ref{nu.z}), the
380: photon sources at smaller red shift, $z \sim z_{\rm min} \approx 5$,
381: are the most important.  But this is only true for the contribution
382: from the most energetic photons, those with energies above the
383: threshold at small red shift, as given by eq.~(\ref{threshold}). As
384: $z$ increases, the threshold energy for muon production decreases, and
385: lower energetic photons can also produce muons.  Moreover, since the
386: photon spectrum is a falling function of energy, low energy photons
387: are more abundant; they all contribute at large red shift.  Which
388: effect wins, which red shift is favored, depends on both the evolution
389: index $m$ and the shape of the photon spectrum.  For a photon spectrum
390: of the form $\dot{n}_\gamma \propto E_\gamma^{-\beta}$ however, the
391: red shift dependence of the flux simplifies.  Using that all photons
392: with energy $E>E_{\rm th}$ produce muons, the neutrino flux then
393: becomes
394: %  
395: \begin{equation}
396: \phi_{\nu} = \xi \dot{n}_0 \frac{3}{2 |\beta - 1|} (10^{20} {\rm
397: eV})^{1-\beta} \,C,
398: \label{nu.beta}
399: \end{equation} 
400: with 	
401: \begin{equation}
402: C = \left\{ 
403: \begin{array}{ll}
404:  	b^{-1} [
405: 	(1 + z_{\rm max} )^{b} - (1 - z_{\rm min} )^{b} ],
406: 	&  \quad b \neq 0, \\ 
407:  	\ln (z_{\rm max}/z_{\rm min}),   
408: 	& \quad b = 0.\\
409: \end{array} \right . 
410: \end{equation}
411: % 
412: Here $\dot{n}_0$ is an overall normalization constant and $b = (2\beta
413: + 3 m -13)/2$.  For $b < 0 $ small red shifts are most important,
414: whereas for $b \ge 0 $ large $z$ dominates.
415: 
416: Nevertheless, the most energetic neutrinos are all produced at small
417: red shift.  We will estimate the flux of these high energy neutrinos,
418: leaving the total flux for the numerical calculation. The highest
419: energy neutrinos are produced at red shift $z \lesssim 10$.  We will
420: assume that the photon spectrum is of the form $\dot{n}_\gamma =
421: \dot{n}_0 E_\gamma^{-\beta} (t/t_0)^m$.  If the photon sources are the
422: origin of the UHECR today one can use the observed UHECR flux to fix
423: the normalization constant $\dot{n}_0$.  This gives
424: %
425: \begin{equation}
426: \phi_\nu = \phi_{_{\rm CR}} \xi \frac{(z_{\rm min})^b - (z_{\rm
427: 	max})^b} {1 - (z_{_{\rm GZk},\gamma})^b},
428: \label{norm}
429: \end{equation} 
430: %
431: where we have used that only photons emitted at red shift $0 < z <
432: z_{_{GZK},\gamma}$ contribute to the observed UHECR flux.  Photons
433: with energies $E_\gamma \sim 10^{20} {\rm eV}$ have an energy
434: attenuation length of $\sim 10 {\rm Mpc}$~\cite{reviews} in the
435: present universe; this corresponds to red shift $z_{_{GZK},\gamma} =
436: 0.002$.  The observed UHECR flux of particles with $E > E_{_{\rm GZK}}
437: \sim 5 \times 10^{19} {\rm eV}$ is $\phi_{_{\rm CR}} \sim 10^{-19}
438: {\rm cm}^{-2} {\rm s}^{-1} {\rm sr}^{-1}$.  Taking $\beta = 2$, in
439: agreement with UHECR observations, and $z_{\rm max} = 10$ we then
440: obtain for the flux of energetic neutrinos:
441: %
442: \begin{equation}
443: \phi_{\nu} \sim \left \{
444: \begin{array}{llll}
445:   7 \times 10^{-20}\, {\rm cm}^{-2} {\rm s}^{-1} {\rm sr}^{-1}, 
446: 	& \qquad m=0, \\
447:   1 \times 10^{-18}\, {\rm cm}^{-2} {\rm s}^{-1} {\rm sr}^{-1}, 
448: 	& \qquad m=1, \\
449:   3 \times 10^{-17}\, {\rm cm}^{-2} {\rm s}^{-1} {\rm sr}^{-1}, 
450: 	& \qquad m=2, \\
451:   7 \times 10^{-16}\, {\rm cm}^{-2} {\rm s}^{-1} {\rm sr}^{-1}, 
452: 	& \qquad m=3. \\
453: \end{array} \right. 
454: \label{flux}
455: \end{equation}
456: %
457: A value of $\beta = 1.5$ or $\beta = 2.5$ changes the produced flux by
458: about a factor $2$. The above results are valid for photon sources
459: that are uniformly distributed throughout the universe. Sources that
460: cluster in galactic halos ({\it e.g.} supermassive particles) produce
461: a lower flux, since the clustering enhances their contribution to the
462: observed UHECR flux.  To get the neutrino flux for photon sources with
463: clustering properties, one has to divide the results in
464: eq.~(\ref{flux}) by the over-density of these sources in our galaxy.
465: 
466: The energy of the produced neutrinos at red shift $z$ is $E_\nu (z)
467: \sim E_\mu/3$.  It is then further red shifted by a factor
468: $(1+z)^{-1}$.  Assuming a falling photon spectrum, we expect most of
469: neutrinos to come from photons near the threshold,
470: eq.~(\ref{threshold}). We estimate the energy of the most energetic
471: neutrinos after the red shift $E_\nu \sim 10^{17}$eV.
472: 
473: The largest flux of neutrinos are produced by $m=3$ sources.  Examples
474: of such sources are necklaces and monopolonium. In general,
475: topological defects emit both UHE protons and photons.  When the
476: protons, and not the photons, are responsible for the UHECR the
477: neutrino flux can no longer be normalized as in
478: eq.~({\ref{norm}). Therefore, whenever we state results for necklaces
479: or monopolonium, we will use the differential photon flux $J(E)$
480: calculated in~\cite{TD_ber,pj,ps} for normalization instead.  For
481: these sources, $\dot{n}_{\gamma,0} (E>E_{\rm min}) \sim L^{-1}
482: \int_{E_{\rm min}} {\rm d}E \, J(E) $, with $L$ the length scale from
483: which the photons are collected.  Because the photon flux is a sharply
484: falling function of energy, $\, \dot{n}_{\gamma,0}$ is dominated by
485: photons with energies $E\sim E_{\rm min}$. Monopolonium clusters in
486: galaxies and has $L \sim L_{\rm gal} \sim 100{\rm kpc}$, the size of
487: our galaxy.  Their over-density in the galaxy is $\sim 2 \times
488: 10^{5}$~\cite{TD_ber}.  Necklaces on the other hand are distributed
489: uniformly throughout the universe; for them $L = L_\gamma \sim 5$Mpc,
490: the photon absorption length at these energies.  We then obtain for
491: the flux of neutrino produced at low red shift:
492: %
493: \begin{equation}
494: \phi_{\nu} \sim \left \{
495: \begin{array}{ll}
496:   10^{-18}{\rm cm}^{-2} {\rm
497:   s}^{-1} {\rm sr}^{-1},  & \quad {\rm monopolonium} \, (m=3) , \\  
498: 10^{-16}{\rm cm}^{-2} {\rm s}^{-1} {\rm
499:   sr}^{-1}, &\quad {\rm necklaces}\, (m=3). \\
500: \end{array} \right. 
501: \label{flux_td}
502: \end{equation}
503: %
504: 
505: 
506: \section{Numerical results}
507: \label{numerics}
508: 
509: \begin{figure}[t]
510: \centering
511: \hspace*{-5.5mm}
512: \leavevmode\epsfysize=6cm \epsfbox{mdiff.eps}\\[3mm]
513: \caption[fig. 1]{\label{fig1} Differential neutrino flux $E^2 J(E)$
514: produced by sources with different evolution indices: $m = 0, \; 1, \;
515: 2, \; 3$ and $4$. For all plots $\beta = 2$, $z_{\rm max} = 10^3$, and
516: $z_{\rm min} = 5$.}
517: \end{figure}
518: 
519: 
520: We have calculated numerically the neutrino flux produced by photon
521: sources at high red shift.  According to our estimate,
522: eq.~(\ref{flux}), $m=3$ sources produce the largest flux of neutrinos,
523: and we will concentrate on them. In particular, we will give a detailed
524: discussion of the neutrino flux produced by necklaces, a candidate
525: $m=3$ source.  Furthermore, we have investigated the possibility that
526: $m=4$ sources, whose density is constrained by bounds on the low
527: energy $\gamma$-ray flux, can nevertheless produce an observable flux
528: of neutrinos.
529: 
530: We studied neutrino production using a Monte Carlo approach.  In our
531: simulation, the red shift at which the UHE photon is emitted and its
532: energy are generated randomly with weight functions $\partial
533: J(E_\gamma, z)/ \partial z$ and $\partial J(E_\gamma, z)/ \partial
534: E_\gamma$ respectively.  Here $J(E_\gamma, z)$ is the differential
535: photon flux generated by the sources under consideration.  The red
536: shift dependence of the photon flux is due to the time evolution of
537: the sources, given by eq.~(\ref{z_dep}), and the expansion of the
538: universe.  The photon spectrum $\partial J(E_\gamma, z)/ \partial
539: E_\gamma$ is a sharply falling function of energy.  It has been
540: calculated for various topological defects and relic particles
541: in~\cite{TD_ber,pj,ps}.
542: 
543: Next, we let the UHE photons scatter with CMB photons, generating the
544: scattering angle and the energy of the background photons randomly for
545: each interaction.  Radio background and intergalactic magnetic fields
546: are insignificant for $z>z_{\rm min}$, and their effects are
547: neglected. The leading particle, i.e. the particle with the highest
548: energy produced in each interaction, is followed throughout the
549: electromagnetic cascade until its energy becomes too low to produce
550: muons.  The probabilities for competing reactions are determined by
551: the cross sections; which reaction actually occurs is once again
552: determined randomly. The reactions taken into account are $\gamma
553: \gamma_{_{CMB}} \rightarrow e^+ e^-$ (PP), $\gamma \gamma_{_{CMB}}
554: \rightarrow e^+ e^- e^+ e^-$ (DPP), $\gamma \gamma_{_{CMB}}
555: \rightarrow \mu^+ \mu^-$ (MP), $e \gamma_{_{CMB}} \rightarrow e \gamma
556: $ (ICS), $e \gamma_{_{CMB}} \rightarrow e\, e^+ e^-$ (TPP) and $e
557: \gamma_{_{CMB}} \rightarrow e\, \mu^+ \mu^-$ (MPP).  The cross section
558: for pion production is small~\cite{pi}, and neglected.
559: 
560: For center of mass energy $s \gg m_e^2$, the cross sections for PP and
561: ICS are related by $\sigma_{_{\rm PP}} = 2 \sigma_{_{\rm ICS}}$, with
562: $\sigma_{_{\rm PP}}$ given by eq.~(\ref{sig_PP}).  The elasticity of
563: these reactions is small: $\eta =1 - 2 (m^2/s)$.  For $s \lesssim 100
564: m^2$, the high energy approximation is not valid, and one has to
565: resort to the exact (tree level) expressions as can be found in any
566: textbook~\cite{peskin}.  The corresponding results for MP can be
567: obtained through the replacement $m_e \leftrightarrow m_\mu$.  The DPP
568: cross section quickly approaches its asymptotic value $\sigma_{\rm
569: DPP} \simeq 6.45 \mu{\rm barn}$.  The kinematics of this reaction has
570: not been calculated.  We have assumed that the energy of the incoming
571: photon is evenly shared by the produced particles, {\it i.e.} four UHE
572: electrons are produced, each with average energy $E_e = E_\gamma/4$.
573: This assumption does not affect the outcome much though: using instead
574: that only two of the electrons get all the energy yields a flux that
575: is almost indistinguishable from the flux obtained by the
576: four-electron-assumption.
577: 
578: The most important interactions for neutrino generation are TPP and
579: MPP: more than $95 \%$ of all cascade interactions are TPP, and MPP
580: governs muon production. For $s \gtrsim 100\, m_e^2$ the cross section
581: for TPP is well approximated by the Borsellino formula~\cite{tpp},
582: which in the high-energy limit reduces to eq.~(\ref{sig_TPP}).  The
583: elasticity of the leading electron is very small, eq~(\ref{eta_tpp}).
584: The cross sections for MPP can be obtained from TPP by replacing $m_e$
585: with $m_\mu$, and multiplying by a symmetry factor $S=2$.  For lower
586: center of mass energies we used the numerical results obtained
587: in~\cite{TD_ber}. If energetic enough, the electron produced together
588: with the muon pair in MPP can produce a pair of muons itself; we have
589: incorporated this effect in our computation.
590: 
591: \begin{figure}[t]
592: \centering
593: \hspace*{-5.5mm}
594: \leavevmode\epsfysize=6cm \epsfbox{bdiff.eps}\\[3mm]
595: \caption[fig. 2]{\label{fig2} Differential neutrino flux $E^2 J(E)$
596: produced by sources with different photon spectra $n_\gamma \propto
597: E_\gamma^{-\beta}$: $\;\beta = 1.5, \; 2$ and $2.5$. For all plots
598: $m=3$, $z_{\rm max} = 10^3$, and $z_{\rm min} = 5$.}
599: \end{figure}
600: 
601: 
602: In figures \ref{fig1} and \ref{fig2} we have plotted the differential
603: neutrino flux produced by various photon sources.  We have assumed
604: that the photon emission rate can be written in the form
605: $\dot{n}_\gamma = \dot{n}_{0,\gamma} E_{\gamma}^{-\beta} (t/t_0)^m$,
606: and have varied the values of $m$ (figure~\ref{fig1}) and $\beta$
607: (figure~\ref{fig2}).  For all plots we have taken $z_{\rm min} =5$,
608: $z_{\rm max} = 10^3$ and photon energies in the range $10^{17} {\rm
609: eV} < E_\gamma < 5 \times 10^{20}$.  To normalize the flux, we have
610: assumed that the photon sources are the origin of the UHECR at
611: present, see eq.~(\ref{norm}).  The results agree well with our
612: estimates.  The highest energy neutrinos have energies $E_\nu \sim
613: 10^{17} {\rm eV}$.  At these energies the differential flux is peaked;
614: for $m=3$ sources the peak value is $E_{\nu}^2 J(E_\nu) = E_{\nu}^2
615: {\rm d} \phi_\nu / {\rm d} E_\nu \sim 10 \, {\rm eV} {\rm cm}^{-2}
616: {\rm s}^{-1} {\rm sr}^{-1}$.  The neutrino flux is smaller for lower
617: values of $m$, and the peak value decreases accordingly: $(E^2 J)_{\rm
618: peak} \sim 1 - 0.1 \, {\rm eV} {\rm cm}^{-2} {\rm s}^{-1} {\rm
619: sr}^{-1} $ for $m=2$ to $(E^2 J)_{\rm peak} \sim 10^{-3}-10^{-2} \,
620: {\rm eV} {\rm cm}^{-2} {\rm s}^{-1} {\rm sr}^{-1}$ for $m=0$.  The
621: flux drops sharply at neutrino energies $E_\nu \sim 10^{18} {\rm eV}$
622: just above the peak energies, due to the cutoff at $z = z_{\rm min} =
623: 5$.  In reality, the radio background and intergalactic magnetic field
624: are turned on gradually, and the cutoff is somewhat less sharp.  The
625: shape of the photon spectrum is not crucial for the results.  Values
626: of $\beta =1.5, \; 2, \; 2.5$ all give similar neutrino spectra, with
627: peak values differing by less than a factor $\sim 2$.
628: 
629: 
630: 
631: \begin{figure}[t]
632: \centering
633: \hspace*{-5.5mm}
634: \leavevmode\epsfysize=6cm \epsfbox{flux.eps}\\[3mm]
635: \caption[fig. 3]{\label{fig3} Neutrino $\phi_\nu$ flux produced by
636: necklaces.  Area under the solid line is the contribution from primary
637: photons to the flux.  Area under the dashed line is the contribution
638: to the flux from red shifts $5 < z < 10$.  }
639: \end{figure}
640: 
641: 
642: We will now discuss in more detail the results for one of the
643: candidate $m=3$ sources, namely necklaces.  The UHE photons emitted by
644: necklaces and other topological defects are the result of decaying
645: superheavy ``X'' particles. For example, X particles may be injected
646: by super conducting strings, emitted from cusps or intersections of
647: ordinary strings, or produced in the annihilation of monopole
648: anti-monopole pair~\cite{TD_review,TD,TD_ber}.  A general feature of
649: decaying supermassive X particles is a predominance of pions among the
650: decay products.  Charged pions decay through the chain $\pi \to \mu
651: \bar{\nu}_\mu \to e \bar{\nu}_e \nu_\mu \bar{\nu}_\mu$, producing
652: electrons and neutrinos.  This gives a flux of primary neutrinos,
653: peaked at high energies: $E_\nu \sim 10^{20} {\rm eV}$.  The photons
654: emitted by neutral pion decay ($\pi^0 \to \gamma \gamma$) together
655: with the electrons from charged pion decay also produce a flux of
656: neutrinos, through the mechanism described in
657: section~\ref{production}.  In our calculation, we have taken into
658: account both the UHE photons and electrons from pion decay.  Since the
659: average electron energy is lower than the average photon energy, the
660: UHE photons will give the dominant contribution to the secondary
661: neutrino flux.
662: 
663: \begin{figure}[t]
664: \centering
665: \hspace*{-5.5mm}
666: \leavevmode\epsfysize=6cm \epsfbox{diff.eps}\\[3mm]
667: \caption[fig. 4]{\label{fig4} Differential neutrino flux $E^2 J(E)$
668: produced by necklaces for various values of $z_{\rm min}$. Here $z3$
669: corresponds to $z_{\rm min} = 3$, $z5$ to $z_{\rm min} = 5$, $z7$ to
670: $z_{\rm min} = 7$, and $z10$ to $z_{\rm min} = 10$.}
671: \end{figure}
672: 
673: 
674: The results for necklaces are shown in figures~\ref{fig3}
675: and~\ref{fig4}. In fig.~\ref{fig3} the neutrino flux as function of
676: neutrino energy is plotted, for $z_{\rm min} =5$, $z_{\rm max} = 10^3$
677: and photon energies $10^{17} {\rm eV} < e_\gamma < 5 \times 10^{20}
678: {\rm eV}$.  We used the photons spectrum calculated (and normalized to
679: fit the UHECR data) in~\cite{TD_ber}, extrapolated down to energies as
680: low as $E_\gamma = 10^{17} {\rm eV}$.  The contribution from the
681: primary photons to the neutrino flux is indicated by the solid line in
682: the plot.  As anticipated, it constitutes the main contribution.  The
683: neutrino flux generated at small red shift $5 < z < 10$ is given by
684: the dashed line. The total flux coming from these red shifts is of the
685: order $\phi_\nu \sim 10^{-16} {\rm cm}^{-2} {\rm s}^{-1}{\rm
686: sr}^{-1}$, in agreement with our estimate, eq.~(\ref{flux_td}).
687: 
688: The neutrino production mechanism is robust, and the produced flux is
689: not much affected by small changes in cross sections or elasticities.
690: Changes of $10\%$ in the MPP cross section, the elasticity of the
691: leading particle and of the produced muons do not change the produced
692: flux by more than a factor $\sim 2$.  The shape of the photon spectrum
693: gives an uncertainty of the same order or magnitude, as can be seen
694: from fig.~\ref{fig2}.  Changing the value of $z_{\rm min}$ makes a
695: more appreciable difference in the produced flux of highest energy
696: neutrinos. Fig.~\ref{fig4} shows $E^2 J(E)$ for values of $z_{\min}
697: =3, \;5, \;7$ and $10$.  Still, the difference in flux with cutoff
698: $z_{\rm min} =3$ and cutoff $z_{\rm min} = 10$ is less than a factor
699: $10$.  For all cutoffs the flux has peak value $E^2 J(E) =
700: \mathcal{O}(10)\; {\rm eV} {\rm cm}^{-2} {\rm s}^{-1} {\rm sr}^{-1}$
701: at the high end of the neutrino spectrum $E_\nu \sim 10^{17} - 10^{18}
702: {\rm eV}$.
703: 
704: 
705: \begin{figure}[t]
706: \centering
707: \hspace*{-5.5mm}
708: \leavevmode\epsfysize=6cm \epsfbox{sdiff.eps}\\[3mm]
709: \caption[fig. 5]{\label{fig5} Maximum differential neutrino flux $E^2
710: J(E)$ that can be produced by $m \ge 4$ sources, in accordance with
711: the EGRET bound.  Line A corresponds to $m = 4$ and $\beta = 1.5$
712: photon spectrum, line B to $m = 4$ and $\beta = 2$, and line C to $m =
713: 5$ and $\beta = 2.5$.}
714: \end{figure}
715: 
716: 
717: 
718: The density of $m \ge 4$ sources, such as superconducting strings, is
719: constrained by the EGRET bound.  The flux of the cascade photons in
720: the energy range $10 \, {\rm MeV} < E <100 \, {\rm GeV}$ must be lower
721: than the flux measured: $\omega_{\rm cas} \le 2.6 \times 10^{-6} {\rm
722: eV} {\rm cm}^3$.  One can estimate the flux of cascade photons
723: produced by TD's by assuming that all the energy density emitted
724: through UHE photons is transferred to the low energy photons.  The
725: universe becomes transparant for low energy photons at red shift $z
726: \sim 10^3$.  The EGRET bound then constrains the photon flux to be
727: %
728: \begin{equation}
729: \int_{z=0}^{z_{_{\rm GZK,\gamma}}} {\rm d} z \; \partial J(E_\gamma,
730:  z) / \partial E_\gamma \lesssim (10^{-2}-10^{-3}) \phi_{_{\rm CR}}.
731: \end{equation}
732: %
733: The maximum flux of neutrinos that can be produced by $m \ge 4$
734: sources when this bound is taken into account is $E^2J(E) \sim
735: \mathcal{O}(1)$, as shown in figure~\ref{fig3}.  Here we used $z_{\rm
736: min} =5$, $z_{\rm max} = 10^3$ and photon energies in the range
737: $10^{17} {\rm eV} < e_\gamma < 5 \times 10^{20} {\rm eV}$.  The photon
738: spectrum is taken to be of the form $\dot{n}_{\gamma, 0} \propto
739: E_\gamma^{-\beta}$. Flux A corresponds to $m=4$ and $\beta = 1.5$,
740: flux B to $m=4$ and $\beta = 2$.  For $m \ge 4$ sources the EGRET
741: constraint is more severe.  Flux C shows the result for $m=5$ and
742: $\beta = 1.5$.
743: 
744:  
745: \section{Discussion}
746: \label{disc}
747: 
748: Neutrinos from slowly decaying relic particles or some other $m \le 2$
749: source are too sparse to be detectable in the foreseeable future.
750: However, the flux from $m=3$ sources, such as necklaces, may be
751: detected soon.  The neutrino spectrum generated by $m=3$ sources has a
752: peak value $(E^2 J)_{\rm peak} = \mathcal{O}(10) \, {\rm eV} {\rm
753: cm}^{-2} {\rm s}^{-1} {\rm sr}^{-1}$ for neutrino energies $E_\nu \sim
754: 10^{17} - 10^{18} {\rm eV}$.  In all our calculations we assumed that
755: the photon sources in question are resposible for the UHECR today.
756: Certainly, the flux of photons emitted cannot exceed the observed
757: UHECR flux, and thus our calculations can be interpreted as an upper
758: bound on the flux of secondary neutrinos produced.
759: 
760: Can other sources produce a comparable flux of neutrinos at
761: $10^{17}-10^{18}$eV?  The neutrino flux $ \phi_{\nu} \sim 10^{-16}
762: {\rm cm}^{-2} {\rm s}^{-1}{\rm sr}^{-1}$ at $E_\nu \sim 10^{17}$eV
763: exceeds the background flux from the atmosphere and from pion
764: photoproduction on CMBR at this energy~\cite{hs,stecker,sp}, as well
765: as the fluxes predicted by a number of
766: models~\cite{whitepaper}. Models of active galactic nuclei (AGN) have
767: predicted a similar flux of neutrinos at these energies~\cite{mpr}.
768: The predictions of these models have been a subject of
769: debate~\cite{wb}.  However, the neutrinos produced by photon sources
770: at high red shift have a distinctive feature: they create a sharp
771: ``bump'' in the spectrum.  Moreover, if the photon sources are TD's
772: the secondary flux is accompanied by a primary neutrino flux peaked at
773: $E_\nu\sim 10^{20}$eV.  And everyone agrees that AGN cannot produce
774: neutrinos with energies of $10^{20}$eV~\cite{rm}.  So, an observation
775: of $10^{17}$eV neutrinos accompanied by a comparable flux of
776: $10^{20}$eV neutrinos would be a signature of a TD rather than an AGN.
777: 
778: The density of TD's with $m \ge 4$, {\em e.g.} superconducting
779: strings, is constrained by the EGRET bound on the flux of low energy
780: $\gamma$-photons.  This also constrains the neutrino flux these
781: sources can produce to be less than $E^2 J(E) \le \mathcal{O}(1) \,
782: {\rm eV} {\rm cm}^{-2} {\rm s}^{-1} {\rm sr}^{-1}$.  This is probably
783: too low to be detected.
784: 
785: To summarize, we have shown that sources of ultrahigh energy photons
786: that operate at red shift $z \gtrsim 5$ produce neutrinos with energy
787: $E_\nu \sim 10^{17}$eV.  The flux depends on the evolution index $m$
788: of the source.  A distinctive characteristic of this type of neutrino
789: background is a ``bump'' in the spectrum at neutrino energies $E_\nu =
790: 10^{17}-10^{18}$eV, and a sharp cutoff at $E_\nu \sim 10^{18}$eV due
791: to the universal radio background at $z<z_{\rm min}$.  We have
792: calculated numerically the neutrino spectrum produced by various
793: sources.  The produced flux is largest for $m=3$ sources. It has a
794: peak value $E^2 J(E) = \mathcal{O}(10)\, {\rm eV} {\rm cm}^{-2} {\rm
795: s}^{-1} {\rm sr}^{-1}$ at neutrino energies $E_\nu \sim 10^{17} -
796: 10^{18} {\rm eV}$.  Detection of these neutrinos can help understand
797: the origin of ultrahigh energy cosmic rays.  Sources with larger
798: values of the evolution index $m \ge 4$ are ruled out as the origin of
799: observed UHECR by the EGRET bound on the flux of $\gamma$-photons.
800: This bound also constrains the flux from these sources to be less than
801: $(E^2 J)_{\rm peak} \le \mathcal{O}(1)\, {\rm eV} {\rm cm}^{-2} {\rm
802: s}^{-1} {\rm sr}^{-1}$.
803: 
804: The author would like to thank A.~Kusenko and C.~Postma for useful
805: discussions. This work was supported in part by the US Department of
806: Energy grant DE-FG03-91ER40662, Task C, and by a grant from UCLA
807: Council on Research.
808:  
809: 
810: 
811: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
812: %% References %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
813: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
814:  
815: \begin{references}
816: 
817: 
818: \bibitem{kp} A.~Kusenko and M.~Postma,
819: %``Neutrinos produced by ultrahigh-energy photons at high red shift,''
820: Phys. Rev. Lett. {\bf 86}, 1430 (2001).  
821: %hep-ph/0007246.
822: %%CITATION = HEP-PH 0007246;%%
823: 
824: \bibitem{data} M.~Takeda {\it et al.}, Phys. Rev. Lett. {\bf 81}, 1163
825: (1998); M.A.~Lawrence, R.J.~Reid and A.A.~Watson, J. Phys. G {\bf G17}, 733
826: (1991); D.~J.~Bird {\it et al.}, Phys. Rev. Lett. {\bf 71}, 3401
827: (1993); Astrophys. J. {\bf 424}, 491 (1994).  
828: 
829: \bibitem{gzk} K.~Greisen, {\sl Phys. Rev. Lett.} {\bf 16}, 748 (1966);
830: G.~T.~Zatsepin and V.~A.~Kuzmin, {\sl Pisma Zh. Eksp. Teor. Fiz.} {\bf
831: 4}, 114 (1966).
832: 
833: \bibitem{reviews} For review, see, {\em e.g.}, P.~L.~Biermann,
834: %``The origin of the highest energy cosmic rays,''
835: J.\ Phys.\ G {\bf G23}, 1 (1997); 
836: %%CITATION = JPHGB,G23,1;%%
837: %\href{\wwwspires?j=JPHGB\%2cG23\%2c1}{SPIRES}
838: P.~Bhattacharjee and G.~Sigl,
839: %``Origin and propagation of extremely high energy cosmic rays,''
840: Phys.\ Rept.\  {\bf 327}, 109 (2000). 
841: %[astro-ph/9811011].
842: %%CITATION = ASTRO-PH 9811011;%%
843: %\href{\wwwspires?eprint=ASTRO-PH/9811011}{SPIRES}
844: 
845: \bibitem{conv} E.~Waxman,
846: %``Cosmological gamma ray bursts and the highest energy cosmic rays,''
847: Phys.\ Rev.\ Lett.\  {\bf 75}, 386 (1995); 
848: %[astro-ph/9505082]; 
849: %%CITATION = ASTRO-PH 9505082;%%
850: %\href{\wwwspires?eprint=ASTRO-PH/9505082}{SPIRES}
851: G.~R.~Farrar and P.~L.~Biermann,
852: %``Correlation between compact radio quasars and ultrahigh energy cosmic
853: %rays,'' 
854: Phys.\ Rev.\ Lett.\  {\bf 81}, 3579 (1998);
855: %[astro-ph/9806242]; 
856: %%CITATION = ASTRO-PH 9806242;%%
857: %\href{\wwwspires?eprint=ASTRO-PH/9806242}{SPIRES}
858: A.~Dar, A.~De Rujula and N.~Antoniou,
859: %``A common origin of all the species of high-energy cosmic rays?,''
860: astro-ph/9901004; 
861: %%CITATION = ASTRO-PH 9901004;%%
862: %\href{\wwwspires?eprint=ASTRO-PH/9901004}{SPIRES}
863: G.~R.~Farrar and T.~Piran,
864: %``GZK violation: A Tempest in a (magnetic) teapot?,''
865: Phys.\ Rev.\ Lett.\  {\bf 84}, 3527 (2000); 
866: %[astro-ph/9906431]; 
867: %%CITATION = ASTRO-PH 9906431;%%
868: %\href{\wwwspires?eprint=ASTRO-PH/9906431}{SPIRES}
869: E.~Ahn, G.~Medina-Tanco, P.~L.~Biermann, and T.~Stanev,
870: %``The origin of the highest energy cosmic rays: Do all roads lead back to
871: %Virgo?,'' 
872: astro-ph/9911123. 
873: %%CITATION = ASTRO-PH 9911123;%%
874: %\href{\wwwspires?eprint=ASTRO-PH/9911123}{SPIRES}
875: 
876: \bibitem{mpr} 
877: K.~Mannheim, R.~J.~Protheroe and J.~P.~Rachen,
878: %``On the cosmic ray bound for models of extragalactic neutrino  production,''
879: Phys. Rev. {\bf D}, to appear [astro-ph/9812398]; astro-ph/9908031. 
880: %%CITATION = ASTRO-PH 9812398;%%
881: 
882: 
883: \bibitem{TD_review} A.~Vilenkin,
884: %``Cosmic Strings And Domain Walls,''
885: Phys.\ Rept.\  {\bf 121}, 263 (1985); 
886: %%CITATION = PRPLC,121,263;%%
887: %\href{\wwwspires?j=PRPLC\%2c121\%2c263}{SPIRES}
888: A.~Vilenkin and E.~P.~S.~Shellard, {\em Cosmic strings and other
889:   topological defects}, Cambridge University Press, Cambridge, England,
890:   1994; 
891: M.~B.~Hindmarsh and T.~W.~Kibble,
892: %``Cosmic strings,''
893: Rept.\ Prog.\ Phys.\  {\bf 58}, 477 (1995). 
894: %[hep-ph/9411342].
895: %%CITATION = HEP-PH 9411342;%%
896: %\href{\wwwspires?eprint=HEP-PH/9411342}{SPIRES}
897: 
898: \bibitem{TD} C.~T.~Hill and D.~N.~Schramm,
899: %``The Ultrahigh-Energy Cosmic Ray Spectrum,''
900: Phys.\ Rev.\  {\bf D31}, 564 (1985); 
901: %%CITATION = PHRVA,D31,564;%%
902: %\href{\wwwspires?j=PHRVA\%2cD31\%2c564}{SPIRES}
903: %
904: C.~T.~Hill, D.~N.~Schramm, and T.~P.~Walker,
905: %``Ultrahigh-Energy Cosmic Rays From Superconducting Cosmic Strings,''
906: Phys.\ Rev.\  {\bf D36}, 1007 (1987);
907: %%CITATION = PHRVA,D36,1007;%%
908: %\href{\wwwspires?j=PHRVA\%2cD36\%2c1007}{SPIRES}
909: %
910: P.~Bhattacharjee, C.~T.~Hill, and D.~N.~Schramm,
911: %``Grand unified theories, topological defects and ultrahigh-energy cosmic
912: %rays,'' 
913: Phys.\ Rev.\ Lett.\  {\bf 69}, 567 (1992).
914: %%CITATION = PRLTA,69,567;%%
915: %\href{\wwwspires?j=PRLTA\%2c69\%2c567}{SPIRES}
916: 
917: \bibitem{TD_ber}
918: V.~Berezinsky, P.~Blasi and A.~Vilenkin,
919: %``Signatures of topological defects,''
920: Phys.\ Rev.\  {\bf D58}, 103515 (1998). 
921: %%CITATION = PHRVA,D58,103515;%%
922: %\href{\wwwspires?j=PHRVA\%2cD58\%2c103515}{SPIRES}
923: 
924: \bibitem{TD_nu}
925: V.~S.~Berezinsky and A.~Vilenkin,
926: %``Ultra high energy neutrinos from hidden-sector topological defects,''
927: hep-ph/9908257.
928: %%CITATION = HEP-PH 9908257;%%
929: 
930: 
931: \bibitem{particles} V.~Berezinsky, M.~Kachelriess, and A.~Vilenkin, Phys.
932: Rev. Lett. {\bf 79}, 4302 (1997); V.~A.~Kuzmin and V.~A.~Rubakov, Phys.
933: Atom. Nucl. {\bf 61}, 1028 (1998) [Yad. Fiz. {\bf 61}, 1122 (1998)];
934: V.~Kuzmin and I.~Tkachev, JETP Lett. {\bf 68}, 271 (1998); M.~Birkel and
935: S.~Sarkar, Astropart. Phys. {\bf 9}, 297 (1998); D.J.~Chung, E.W.~Kolb, and
936: A.~Riotto, Phys. Rev. Lett.  {\bf 81}, 4048 (1998); Phys. Rev. {\bf D59},
937: 023501 (1999); K.~Benakli, J.~Ellis, and D.~V.~Nanopoulos, Phys. Rev. {\bf
938: D59}, 047301 (1999); Phys.  Rev. {\bf D59}, 123006 (1999).
939: 
940: \bibitem{gk2}
941: G.~Gelmini and A.~Kusenko,
942: %``Unstable superheavy relic particles as a source of neutrinos
943: %responsible for the ultrahigh-energy cosmic rays,'' 
944: Phys.\ Rev.\ Lett.\  {\bf 84}, 1378 (2000). 
945: %[hep-ph/9908276].
946: %%CITATION = HEP-PH 9908276;%%
947: %\href{\wwwspires?eprint=HEP-PH/9908276}{SPIRES}
948: J.~L.~Crooks, J.~O.~Dunn, and P.~H.~Frampton,
949: %``Relic Neutrinos and Z-Resonance Mechanism for Highest-Energy Cosmic Rays,''
950: astro-ph/0002089.
951: %%CITATION = ASTRO-PH 0002089;%%
952: %\href{\wwwspires?eprint=ASTRO-PH/0002089}{SPIRES}
953: 
954: \bibitem{kt} For review, see, {\em e.g.}, 
955: V.~A.~Kuzmin and I.~I.~Tkachev,
956: %``Ultra high energy cosmic rays and inflation relics,''
957: Phys.\ Rept.\  {\bf 320}, 199 (1999). 
958: %[hep-ph/9903542].
959: %%CITATION = HEP-PH 9903542;%%
960: %\href{\wwwspires?eprint=HEP-PH/9903542}{SPIRES}
961: 
962: 
963: 
964: \bibitem{hs}
965: C.~T.~Hill and D.~N.~Schramm,
966: %``Ultrahigh-energy cosmic ray neutrinos,''
967: Phys.\ Lett.\  {\bf B131}, 247 (1983).
968: %%CITATION = PHLTA,B131,247;%%
969: %\href{\wwwspires?j=PHLTA\%2cB131\%2c247}{SPIRES}
970: 
971: \bibitem{stecker} F.~W.~Stecker, C.~Done, M.~H.~Salamon, and P.~Sommers,
972: %``High-energy neutrinos from active galactic nuclei,''
973: Phys.\ Rev.\ Lett.\ {\bf 66}, 2697 (1991); erratum: {\em ibid}., {\bf 69}, 
974: 2738 (1992).
975: %%CITATION = PRLTA,66,2697;%%
976: %\href{\wwwspires?j=PRLTA\%2c66\%2c2697}{SPIRES}
977: 
978: \bibitem{ysl} S.~Yoshida, G.~Sigl, and S.~Lee,
979: %``Extremely high energy neutrinos, neutrino hot dark matter, and the
980: %highest energy cosmic rays,'' 
981: Phys.\ Rev.\ Lett.\  {\bf 81}, 5505 (1998).
982: %[hep-ph/9808324].
983: %%CITATION = HEP-PH 9808324;%%
984: %\href{\wwwspires?eprint=HEP-PH/9808324}{SPIRES}
985: G.~Sigl, S.~Lee, P.~Bhattacharjee, and S.~Yoshida,
986: %``Probing grand unified theories with cosmic ray, gamma-ray and neutrino
987: %astrophysics,'' 
988: Phys.\ Rev.\  {\bf D59}, 043504 (1999)
989: %[hep-ph/9809242].
990: %%CITATION = HEP-PH 9809242;%%
991: %\href{\wwwspires?eprint=HEP-PH/9809242}{SPIRES}
992: 
993: \bibitem{pj} 
994: R.~J.~Protheroe and P.~A.~Johnson,
995: %``Propagation of ultrahigh-energy protons over cosmological distances and
996: %implications for topological defect models,'' 
997: Astropart.\ Phys.\  {\bf 4}, 253 (1996)
998: [astro-ph/9506119].
999: %%CITATION = ASTRO-PH 9506119;%%
1000: 
1001: \bibitem{ps} R.~J.~Protheroe and T.~Stanev,
1002: %``Limits on models of the ultrahigh energy cosmic rays based on
1003: %topological defects,'' 
1004: Phys.\ Rev.\ Lett.\  {\bf 77}, 3708 (1996)
1005: [astro-ph/9605036].
1006: %%CITATION = ASTRO-PH 9605036;%%
1007: 
1008: \bibitem{whitepaper} D.~B.~Cline and F.~W.~Stecker,
1009: %``Exploring the ultrahigh energy neutrino universe,''
1010: astro-ph/0003459.
1011: %%CITATION = ASTRO-PH 0003459;%%
1012: %\href{\wwwspires?eprint=ASTRO-PH/0003459}{SPIRES}
1013: 
1014: 
1015: \bibitem{berezinsky} V.~Berezinsky, Sov. J. of Nucl. Phys. {\bf 11}, 222
1016: (1970).
1017: 
1018: 
1019: \bibitem{biermann}
1020: R.~J.~Protheroe and P.~L.~Biermann,
1021: %``A new estimate of the extragalactic radio background and implications
1022: %for ultrahigh-energy gamma ray propagation,'' 
1023: Astropart.\ Phys.\  {\bf 6}, 45 (1996).
1024: %[astro-ph/9605119].
1025: %%CITATION = ASTRO-PH 9605119;%%
1026: %\href{\wwwspires?eprint=ASTRO-PH/9605119}{SPIRES}
1027: 
1028: \bibitem{condon}J.~J.~Condon, Astrophys. J. {\bf 284}, 44 (1984).
1029: 
1030: \bibitem{recent} P.~Madau et al., Mon. Not. R. Astron. Soc. {\bf 283}, 35
1031: (1996); C.C.~Steidel, Proc. Nat. Acad. Sci. {\bf 96}, 4232, 1999;
1032: T.~Miyaji, G.~Hasinger, and M.~Schmidt, astro-ph/9809398; 
1033: G. Pugliese, H. Falcke, Y. Wang, and P.L. Biermann, Astron. and Astrophys.,  
1034: {\bf 358}, 409 (2000).
1035: %%CITATION = ASTRO-PH 0003025;%%
1036: 
1037: \bibitem{pi}
1038: J.~Boyer {\it et al.},
1039: %Two photon production of pion pairs
1040: Phys.\ Rev.\ {\bf D42}, 1350 (1990).
1041: 
1042: \bibitem{dpp} 
1043: R.~W.~Brown, W.~F.~Hunt, K.~O.~Mikaelian and
1044: I.~J.~Muzinich, Phys.\ Rev.\ {\bf D8}, 3083 (1973).
1045: 
1046: 
1047: \bibitem{tpp}
1048: A.~Borsellino, Nuovo Cimento {\bf 4}, 112 (1947); 
1049: K.~J.~Mork
1050: %``Pair production by photons on electrons,''
1051: Phys.\ Rev.\ {\bf 160}, 1065 (1967);
1052: A.~Mastichiadis, A.~P.~Marscher and K.~Brecher. 
1053: 
1054: \bibitem{tpp2}
1055: A.~Mastichiadis,  A.~P.~Marscher, and K.~Brecher, 
1056: %``Electron-positron pair production by ultrarelativistic eleectrons 
1057: % in a soft photon field,''
1058: Astrophys. Journal {\bf 300}, 178 (1986);
1059: V.~Anguelov, S.~Petrov, L.~Gurdev and J.~Kourtev,
1060: %``On the numerical analysis of triplet pair production cross-sections
1061: % and the mean eneergy of produced particles for modelling electron-photon
1062: %cascade in a soft photon field,''
1063: J.\ Phys.\ {\bf G25}, 1733 (1999).
1064: %[astro-ph/0001252]
1065: 
1066: 
1067: \bibitem{ggs} 
1068: P.~Gondolo, G.~Gelmini and S.~Sarkar,
1069: %``Cosmic neutrinos from unstable relic particles,''
1070: Nucl.\ Phys.\  {\bf B392}, 111 (1993).
1071: %[hep-ph/9209236].
1072: %%CITATION = HEP-PH 9209236;%%
1073: %\href{\wwwspires?eprint=HEP-PH/9209236}{SPIRES}
1074: 
1075: \bibitem{peskin} M.~E.~Peskin and D.~V.~Schroeder, {\em An introduction to
1076: quantum field theory}, Addison-Wesley Publishing Company, 1995.
1077: 
1078: \bibitem{egret} EGRET collaboration, P.~Streekumar {\it et al.},
1079: Astrophys.~J. {\bf 494}, 523 (1998).
1080: 
1081: 
1082: \bibitem{sp}
1083: T.~Stanev, R.~Engel, A.~Muecke, R.~J.~Protheroe and J.~P.~Rachen,
1084: %``Propagation of ultrahigh energy protons in the nearby universe,''
1085: Phys.\ Rev.\  {\bf D62}, 093005 (2000)
1086: [astro-ph/0003484].
1087: %%CITATION = ASTRO-PH 0003484;%%
1088: 
1089: \bibitem{wb}
1090: E.~Waxman and J.~Bahcall,
1091: %``High energy neutrinos from astrophysical sources: An upper bound,''
1092: Phys.\ Rev.\  {\bf D59}, 023002 (1999).
1093: %%CITATION = HEP-PH 9807282;%%
1094: %\href{\wwwspires?eprint=HEP-PH/9807282}{SPIRES}
1095: 
1096: 
1097: \bibitem{rm} 
1098: J.~P.~Rachen and P.~M\'esz\'aros,
1099: %``Photohadronic neutrinos from transients in astrophysical sources,''
1100: Phys.\ Rev.\  {\bf D58}, 123005 (1998)
1101: [astro-ph/9802280].
1102: %%CITATION = ASTRO-PH 9802280;%%
1103: 
1104: 
1105: \end{references}
1106: 
1107: 
1108: \end{document}
1109: 
1110: 
1111: 
1112: 
1113: