1: \documentclass[12pt,preprint]{aastex}
2: %\usepackage{amsmath, amsthm, aastex}
3:
4: %\usepackage{amsmath, amsthm, lscape}
5: %\usepackage{lscape}
6: \shorttitle{Cluster Neutrinos}
7: \shortauthors{Wolfe}
8:
9: \begin{document}
10:
11: \title{Neutrinos and Gamma Rays from Galaxy Clusters}
12:
13: \author{Brandon Wolfe\altaffilmark{1,5}, Fulvio Melia\altaffilmark{1,2}, Roland M. Crocker\altaffilmark{3,4},
14: and Raymond R. Volkas\altaffilmark{5}}
15: \altaffiltext{1}{Physics Department, The University of Arizona, Tucson, AZ 85721}
16: \altaffiltext{2}{Steward Observatory, The University of Arizona, Tucson, AZ 85721}
17: \altaffiltext{3}{School of Chemistry and Physics, The University of Adelaide, South Australia 5005}
18: \altaffiltext{4}{Current address: J.L. William Fellow, School of Physics, Monash University,
19: Victoria 3800, Australia}
20: \altaffiltext{5}{School of Physics, Research Centre for High Energy Physics, University of
21: Melbourne, Victoria 3010, Australia}
22:
23: \begin{abstract}
24: The next generation of neutrino and $\gamma$-ray detectors should provide new insights
25: into the creation and propagation of high-energy protons within galaxy clusters,
26: probing both the particle physics of cosmic rays interacting with the background medium
27: and the mechanisms for high-energy particle production within the cluster.
28: In this paper we examine the possible detection of $\gamma$-rays (via the GLAST satellite)
29: and neutrinos (via the ICECUBE and Auger experiments) from the Coma cluster of galaxies,
30: as well as for the $\gamma$-ray bright clusters Abell 85, 1758, and 1914. These three
31: were selected from their possible association with unidentified EGRET sources, so it
32: is not yet entirely certain that their $\gamma$-rays are indeed produced diffusively within
33: the intracluster medium, as opposed to AGNs. It is not obvious why these inconspicuous
34: Abell-clusters should be the first to be seen in $\gamma$-rays, but a possible reason
35: is that all of them show direct evidence of recent or ongoing mergers. Their identification
36: with the EGRET $\gamma$-ray sources is also supported by the close correlation between
37: their radio and (purported) $\gamma$-ray fluxes. Under favorable conditions (including
38: a proton spectral index of $2.5$ in the case of Abell 85, and $\sim 2.3$ for Coma,
39: and Abell 1758 and 1914), we expect
40: ICECUBE to make as many as $0.3$ neutrino detections per year from the Coma cluster of
41: galaxies, and as many as a few per year from the Abell clusters 85, 1758, and 1914.
42: Also, Auger may detect as many as 2 events per decade at $\sim 10^{18}$ eV from
43: these gamma-ray bright clusters.
44: \end{abstract}
45:
46: \keywords{acceleration of particles --- galaxies: clusters: individual (Coma) ---
47: neutrinos --- radiation mechanisms: non-thermal --- relativity --- X-rays: galaxies}
48:
49: \section{Introduction}
50:
51: A population of high-energy cosmic-ray (CR) protons, pervading galaxy clusters, was
52: first invoked (Dennisson 1980) to explain highly polarized radio luminosities ($L_r
53: \sim 10^{40}-10^{42}$ erg s$^{-1}$) now observed in over 30 clusters. Evidently the
54: radio light is synchrotron emission from highly relativistic electrons ($\gamma_e
55: \sim 10^6$) gyrating in large magnetic fields ($B \sim 1 \ \mu$G), despite the fact
56: that such electrons would lose their energy to radiation well before they could
57: traverse the $\sim$Mpc extent of observed emission. However, protons injected into
58: the cluster diffuse throughout the intracluster medium, essentially unchanged and
59: confined for cosmological times. In this environment they may interact with the
60: background gas and create a decay cascade. Among the products of this decay are
61: the relativistic electrons responsible for radio emission.
62:
63: Since Dennisson's (1980) pioneering work, conclusive observations of this `secondary'
64: process for electron production within clusters have been few. In Hercules A (Nulsen
65: et al. 2005b), MS 0735 (McNamara et al. 2005), and Hydra A (Nulsen et al. 2005b),
66: vacated areas of X-ray bremsstrahlung indicate the background of thermal electrons
67: is being pushed aside by relativistic electrons injected from the axes of a central
68: source, implying a phenomenally large power, $L_p \sim 10^{44}$ erg s$^{-1}$. These
69: `bubbles' coincide with sources of radio emission as the protons produce a cascade
70: of synchrotron-emitting electrons, confirming that the secondary decay process does
71: occur in clusters. The protons released during such an episode would be confined and
72: built up within the cluster over cosmic times, totaling some $10^{61}$ erg of nonthermal
73: hadrons---very near energy equipartition with the thermal background (Berezinsky, Blasi,
74: and Ptuskin 1997; Hinton \& Domainko 2006). This is, we find, the order of nonthermal
75: energy required to be deposited in the nonthermal hadron population for proton-proton
76: scatterings to produce $\pi^0$ decay $\gamma$-rays within the EGRET detection sensitivity.
77:
78: The energy liberated by the infall of the galaxy group NGC~4839 toward the center of
79: the Coma cluster, observed by XMM-Newton (Neumann et al. 2001), is an example of another
80: potential source of high-energy protons. NGC~4839 achieves a velocity of $\sim 1,400$ km
81: s$^{-1}$, and since the sound speed corresponding to Coma's gas temperature of $\sim 8$
82: keV is $\sim 1,000$ km s$^{-1}$, the subcluster's supersonic motion is expected to
83: produce shocks (which Neumann et al. claim to observe directly in the imaging of
84: NGC~4839). It is likely that the shock compression in these regions energizes a fraction
85: of the thermal particles within the intracluster medium (ICM) by first-order Fermi
86: acceleration. However, how the protons are actually energized is not directly relevant
87: to our analysis. For example, since the cosmic rays remain trapped within the cluster
88: for over a Hubble time, it is likely that second-order acceleration by a turbulent
89: distribution of Alfv\'en waves may be even more effective at helping the protons attain
90: their highest energies (see, e.g., Liu, Petrosian, and Melia 2004; Liu, Melia, and
91: Petrosian 2006). Our principal goal is to determine the observational signatures
92: such a population would produce, should it be present within the cluster. But cluster
93: mergers do provide us with a viable mechanism for energizing the cosmic rays, regardless
94: of how the individual particles are ultimately accelerated. In this regard, we note that
95: the gravitational energy $\sim 10^{64}$ erg released in a large relic merger could
96: reasonably supply $\sim 10^{62}$ erg in a nonthermal hadron population, again
97: sufficient to supply detectable $\gamma$-rays. In this case, the radio halo may
98: be interpreted as a signature of the cosmological development of a cluster (the
99: oldest known cluster being some 9 billion years old).
100:
101: Aside from a radio halo produced by secondary electrons,
102: conclusive evidence for the properties of CR protons within galaxy clusters
103: rests on the detection of other products of the protons' decay cascade: $\gamma$-rays
104: and neutrinos. The closest active galaxy cluster, the Coma cluster, does not emit
105: copiously enough to have been detected by EGRET given its $\sim10^{-11}$ cm$^{-2}$
106: s$^{-1}$ flux sensitivity, although we will show that this does not
107: preclude a possible detection by the GLAST satellite or a
108: detectable neutrino flux. Our preliminary identification of clusters which \emph{are}
109: $\gamma$-ray bright, and, therefore, plausible extrasolar neutrino candidates,
110: has come from a systematic search for likely clusters among unidentified EGRET sources.
111:
112: In a search for new radio galaxy clusters, the NRAO VLA (Giovannini et al. 1999) and
113: Westerbork Northern (Kempner \& Sarazin 2001) sky surveys were systematically cross
114: referenced with X-ray bright objects from the ROSAT survey (Ebeling et al.
115: 1998), providing a list of more than 30 known cluster halos or relics,
116: presented in Table 1. While none of the clusters EGRET was hoped to detect
117: produce a measurable $\gamma$-ray flux (Reimer et al. 2003), three radio
118: sources coincide with unidentified $\gamma$-ray bright objects detected by EGRET
119: (Colafrancesco 2002)---the cluster Abell 85 and the EGRET source 3EG J0038-0949,
120: Abell 1914 and EGRET source J1424+3734, and Abell 1758 and EGRET source 3EG J1337 +502a.
121: These three EGRET sources are not unequivocally the quoted clusters, but here we
122: shall assume they are in fact the same objects.
123:
124: The assumption is not certain. In a study by Reimer et al. (2003) it was shown that
125: the probability that one of the 170 unidentified EGRET sources and a member of
126: their list of 58 considered galaxy clusters coincided was 48.1\%.
127: Of their list of 58 clusters worthy of consideration, only
128: Abell 85 and the unidentified source J0038-0949 showed a considerable overlap,
129: which could therefore be explained as an entirely chance occurrence.
130: However only six of their 58 considered clusters
131: ---included for their proximity ($z<0.14$) and X-ray brightness---
132: were shown by the NRAO VLA or Westerbork Northern
133: radio sky surveys to be radio halo/relic clusters, so that in only these six
134: cases is there any reason to believe a significant nonthermal proton population may be present.
135: Also, Abell 1758 and 1914 were not considered by Reimer et al. (2003), as they are above $z >0.14$.
136:
137: The three candidates are not the most X-ray bright clusters, most of which were
138: assigned only upper limits in the nine years of EGRET observations. It is not
139: obvious why three inconspicuous Abell-clusters are the first to be seen in gamma
140: rays when the closest and most well-studied clusters---known to have evidence for
141: non-thermal particle populations from radio, EUV, or hard X-ray detections---are not.
142: Perhaps a key is that
143: each of the $\gamma$-ray bright candidates shows direct evidence of recent or
144: ongoing mergers, indicating (if confirmed) that this may outcompete the AGN outburst
145: as a source of cluster cosmic rays. In Abell 85, a 4 Mpc filament (Durret
146: et al. 2005) forms a chain of several groups of galaxies.
147: In particular, an impact region (the `south blob') displays a temperature ($8.8$ keV) far
148: in excess of the rest of the cluster ($2.7$ keV). In Abell 1914 (Govoni et al. 2004)
149: a NE-SW arch-like hot region extends across the cluster
150: center and may represent an ongoing shock. Abell 1758 is a double cluster
151: (David \& Kempner 2004): A1758N is in the late stages of a large impact
152: parameter merger between two $7$ keV clusters, while A1758S is
153: undergoing a smaller impact parameter merger between two 5 keV cores.
154: In each case, the flux of $\gamma$-rays correlates with the radio flux, as it would if
155: both were linked to secondary decay products. Abell 85 also shows evidence of a
156: hard X-ray excess (Lima-Neto et al. 2001), indicative of
157: a nonthermal population of electrons.
158:
159: In the end, however, we caution that while the identification of Abell clusters
160: 85, 1758, and 1914 with three unidentified EGRET sources may be motivated by the
161: reasons we have outlined here, it is not entirely clear that their $\gamma$-rays
162: are indeed produced diffusively within the ICM. Thus, one must accept the results
163: presented in this paper, particularly the predicted neutrino fluxes, with this
164: important caveat in mind.
165:
166: As clusters confine CR protons over cosmological times, they provide a promising
167: source for a possible first detection of high-energy ($\gtrsim$ TeV) extra-solar
168: neutrinos. Aside from probing high or ultra-high energy particle physics processes
169: occurring within astrophysical sources, any such neutrino (or $\gamma$-ray) detection
170: would also establish the properties of the proton population within the cluster
171: and settle the possible role of such populations in various cluster
172: phenomena---e.g., generating the radio halo (Dennisson 1980), their contribution to
173: the `cooling flow' problem (Silk 1995), or their acceleration within cluster
174: merger shocks (Sarazin 2004).
175:
176: In this paper we take the optimistic view that protons within the Coma cluster
177: create a flux of $\gamma$-rays which is within the detection limits
178: of the GLAST satellite, without violating the EGRET null detection.
179: In the same spirit of optimism, we identify the Abell clusters 85, 1758, and 1914 with
180: their gamma-ray bright counterparts. The resulting predictions for neutrino
181: and gamma-ray detections for these four clusters will determine whether this optimism was
182: well-founded.
183:
184: Work on the hadronic origin of one or more spectral components in the overall
185: cluster emissivity has been quite extensive, and our investigation overlaps with several
186: earlier treatments, though our model is unique in the treatment of the particle
187: kinetics and proton-electron scattering events. Dolag and Ensslin (2000) considered
188: the possible origin of the radio halos from a hadronic secondary electron injection,
189: based on a detailed modeling of the cluster magnetic field (see also Blasi \&
190: Colafrancesco 2000, Jones 2004, Brunetti et al. 2007, Marchegiani et al. 2007).
191: Subsequently, Pfrommer and Ensslin (2004) constrained the cosmic-ray population
192: in several cooling flow clusters from their radio and $\gamma$-ray emissivities (see
193: also Reimer et al. 2004, and Sarazin 2007). Not surprisingly, the $\gamma$-ray emission
194: from proton-proton collisions in clusters may contribute a non-negligible fraction
195: of the extragalactic background radiation field (Kuo et al. 2005). And though
196: we will not be addressing $\gamma$-ray line emission in this paper, cosmic rays may
197: also collide with heavier nuclei in the ICM to produce measurable $\gamma$-ray
198: line intensities (Iyudin et al. 2004). An extensive review of the $\gamma$-ray
199: emissivity in clusters may be found in Blasi et al. (2007).
200:
201: The issue of hadronic acceleration during large-scale structure formation has been
202: considered by several authors, including Miniati et al. (2001), Berrington \& Dermer
203: (2003), Blasi (2004), Ensslin et al. (2007), Jubelgas et al. (2008), Ando and Nagai
204: (2008), Nakar et al. (2008) and, in the context of secondary acceleration due to
205: stochastic processes, Brunetti et al. (2004), Brunetti and Blasi (2005), and Petrosian
206: and Bykov (2008). But cosmic rays may also play an important dynamical role throughout
207: the cluster's lifetime. Collisions between energetic hadrons and the ICM have been
208: invoked as a means of heating the gas condensing toward the middle of the cluster,
209: thereby preventing an overly rapid cooling of the flow (Colafrancesco, Dar, and
210: De R\'ujula 2004, Guo \& Oh 2008).
211:
212: Our goal in this paper is specifically to calculate the neutrino flux expected
213: from $\gamma$-ray bright clusters, but our approach differs from earlier attempts
214: in several important ways, which we will describe over the next several sections.
215: Our model requires that the number of protons
216: be consistent with a secondary explanation of the Coma cluster's radio halo, using
217: magnetic fields ($\sim 1 \mu$G) consistent with observations of Faraday rotation,
218: rather than those ($\sim 0.1 \mu$G) consistent with an inverse-Compton description
219: of the hard X-ray excess detected by Beppo\emph{SAX} and RXTE (see also Wolfe and
220: Melia 2006). The only model-dependent characteristics of this work are then the
221: distribution of gas and that of protons diffusing throughout the cluster, which---together
222: with the details of calculating the various links in the decay cascade---we discuss in
223: Section 2. Our technique for calculating the detected number of neutrinos, given the predicted flux
224: of Coma, is discussed in Section 3. Finally in Section 4, we review the forecast for
225: those clusters for which a positive $\gamma$-ray detection has been proposed.
226:
227: \section{Model Characteristics and Flux Calculation}
228: \subsection{Emission spectra of nonthermal plasmas}
229: Our model first requires the distribution of electrons, neutrinos, and $\gamma$-rays
230: produced by proton-proton collisions and subsequent decay of secondaries.
231: Gamma rays are produced from the decay of neutral
232: pions $pp \rightarrow \pi^0 \rightarrow 2 \gamma$ at a
233: characteristic energy of $\sim$ 70 MeV in the pion rest frame;
234: a similar number of neutrinos are also produced via
235: $pp \rightarrow \pi^\pm \rightarrow \mu^\pm + \nu_\mu(\bar{\nu}_\mu)$,
236: and in the subsequent decay of muons into electrons and positrons, $\mu^\pm \rightarrow
237: e^\pm + \bar{\nu}_\mu (\nu_\mu) + \nu_e (\bar{\nu}_e)$ (given
238: typical power-law parent proton distributions,
239: neutrinos from neutron $\beta$-decay
240: may be neglected in these calculations). Both neutrinos and
241: $\gamma$-rays may also be produced via proton interactions with background
242: light fields. For the cosmic microwave background, pion production is characterized
243: by a peak at $6 \times 10^{19}$ eV---approximately the energy of a 140 km/h tennis
244: ball---while for infrared light emitted by galaxies and proposed
245: (Blasi 2005) as a dominant source, this peak shifts to $2.5 \times 10^{18}$ eV.
246: All three decay sources probe different energy regimes: electrons, via synchrotron radio
247: ($10^{-6}$ eV) or HXR inverse-Compton ($10$ eV); $\gamma$-rays, at 70 MeV
248: and notably detectable by GLAST and EGRET; and TeV+ neutrinos, which directly probe the
249: tail of the proton distribution ($10^{13}$--$10^{15}$ eV) via pp scatterings as well
250: as its extreme limit ($>10^{18}$ eV) via interactions with the CMB.
251:
252: Our recipe is to assume or imply (via radio observations, see Sec 3.2) a proton
253: distribution, which in a practical sense means setting a spectral index and
254: normalization to a power-law $n_p(E_p) = n_{p0} E_p^{s_p}$. From this alone follow
255: two injections of pions---one for proton interactions (pp) with hydrogen in
256: the ICM, and one for interactions with background light (p$\gamma$). The
257: observationally relevant decay products of these pions are electrons,
258: $\gamma$-rays, and neutrinos (see also Blasi and Colafrancesco 1998), which
259: for the particles of interest may be
260: found by summing over the kinetically allowed energies in each link of a
261: decay chain; this section is devoted to a description of this process.
262:
263: For proton-proton interactions, the neutral pion emissivity is
264: \begin{equation}
265: q^{pp}_{\pi^0}(E_{\pi^0}) = c n_H \int_{E_{thres}} dE_p n_p(E_p) \frac{d \sigma(E_{\pi^0},E_p)}{d E_{\pi^0}}\;,
266: \end{equation}
267: and similarly for charged pions, via an altered cross section.
268: The threshold proton kinetic energy for production of a pion in a proton-proton collision
269: is $E_{thres} = 2 m_\pi (1 + m_\pi/4 m_p)$.
270:
271: Here we take the cross-section
272: for pion production for energies below the isobar resonance ($E_p = 7$ GeV) from
273: Dermer (1986)---where also each of the various decay channels is treated---while
274: above the resonance we take the cross-section to be a simple scaling form
275: used by Blasi $\&$ Colafrancesco (1999), and Fatuzzo $\&$ Melia (2003).
276: We note in this context that while collider data at these extreme energies
277: do not exist, the proton-proton collision cross-section has
278: been examined by the Fly's Eye detector, which implies from atmospheric
279: fluorescence that the collision cross-section violates the
280: scaling assumption when $\sqrt{s}$ is tens of TeV.
281: Unfortunately the pion production cross-section is determined both
282: by the rate of proton collisions, $R_{pp} = n_H c\, d\sigma_{pp}/dE_p$,
283: and by the varying multiplicity, $M(E_\pi,E_p)$, of pion production with proton energy
284: (Moffeit et al. 1972, Markoff and Melia 1997, 1999). For this we again
285: depend on accelerator data, which exist only up to 1,800 GeV---so that the
286: limitation on our knowledge of the cross-section is compounded by a lack of
287: certainty regarding pion production. Simply, predictions and observations of
288: neutrinos or $\gamma$-rays in clusters probe energy regimes not previously accessible.
289:
290: Once the pion source function is found, the neutrino, $\gamma$-ray, and electron fluxes
291: follow from kinematic concerns discussed fully in Stecker (1979) for $\gamma$-ray decays
292: and Marscher et al. (1980), and Zatsepin $\&$ Kuz'min (1962) for neutrinos.
293: For decay photons we have
294: \begin{equation}
295: q^{pp}_{\gamma}(E_\gamma) = 2 \int_\phi dE_{\pi^0} \frac{ q^{pp}_{\pi^0}(E_{\pi^0}) }
296: {(E_{\pi^0}^2 - m_{\pi^0}^2 c^4)^{1/2}}\;,
297: \end{equation}
298: where the minimum kinetically allowed pion energy is $\phi = E_\gamma + m^2_{\pi^0}
299: c^4/(4 E_\gamma)$. Well above the 70 MeV region,
300: the pion spectrum shares the proton parents' spectral index, $s_\pi = s_p$.
301: For electrons, the source function is
302: \begin{equation}
303: q^{pp}_e(E_e) = n_H c \frac{m_\pi^2}{m_\pi^2 - m_\mu^2}
304: \int^{E_p^{max}}_{E_e}
305: dE_\mu \frac{dP}{dE_e}
306: \int^{E_\pi^{max}}_{E_\pi^{min}}
307: \frac{dE_\pi}{\beta_\pi E_\pi}
308: \int^{E_p^{max}}_{E_{thres}(E_\pi)}
309: dE_p n_p(E_p) \frac{d\sigma(E_\pi,E_p)}{dE_\pi}\;,
310: \end{equation}
311: where $dP/dE_e$ is the three-body decay probability (Fatuzzo \& Melia 2003). Here, the limits
312: of kinematically allowed pion energies are
313: \begin{equation}
314: E_{\pi}^{min,max} = \frac{2 E_\mu}{(1\pm\beta_\mu)+m_\pi^2/m_\mu^2(1\mp\beta_\mu)}
315: \end{equation}
316: (with minimum corresponding to the upper sign, maximum to the lower).
317: The threshold $E_{thresh}(E_\pi)$for producing a pion with energy $E_\pi$ is similar to that in Eq. (1).
318:
319: For neutrinos, we must distinguish between particles produced by pion decay, and those
320: produced by muon decay, since these are two- and three-body decays, respectively.
321: For pions the source function is
322: \begin{equation}
323: q^{pp}_{\nu_\alpha}(E_{\nu_\alpha}) = \int_{\gamma^\pi_{min}}^\infty
324: f(\gamma_\pi) q_\pi(\gamma_\pi) d\gamma_\pi\;.
325: \end{equation}
326: In the case of electron injection resulting from proton collisions we integrated over the three-body
327: decay probability $dP/dE_e$. Here $f(\gamma_\pi)$ plays that role, as the spectrum of two-body decay
328: \begin{equation}
329: f(\gamma_\pi) = \frac{1}{2 E_\nu^0 \sqrt{\gamma_\pi^2 - 1}}\;.
330: \end{equation}
331: Here, $E_\nu^0 \approx 29.8$ MeV is the CMS energy of the emitted neutrino. In this
332: instance the minimum kinetically allowed energy is
333: \begin{equation}
334: \gamma_{min}^\pi = E_\nu^0/2E_\nu + E_\nu/2E_\nu^0\;.
335: \end{equation}
336:
337: In the decay of charged muons, the muon neutrino flux is
338: \begin{equation}
339: q_\nu^\mu (E_\nu) = \frac{1}{2} \int_{\gamma_\mu,min}^\infty f(E_\nu) \int^1_{\chi_{min}} q_\pi(\gamma_\pi)
340: \frac{\partial \gamma_\pi}{\partial \gamma_\mu}\; d\chi \;d\gamma_\mu\;,
341: \end{equation}
342: where electron neutrinos are produced at approximately half the rate of these muon neutrinos.
343: The interaction angle cosine, $\chi$, is kinematically limited to
344: \begin{eqnarray}
345: \chi_{min}&=&-1\,,\quad{\rm for}\; 3.68 \le 3.55 \gamma_\mu\nonumber\\
346: &=&\sqrt{13.54 - 12.6 \gamma_\mu^2}\,,\quad{\rm for}\; 3.68 > 3.55 \gamma_\mu\;,
347: \end{eqnarray}
348: while the lower limit for $\gamma_\mu$ is
349: \begin{equation}
350: \gamma_{\mu,min} = {E_\nu\over m_\mu} + {m_\mu\over 4 E_\nu}\;.
351: \end{equation}
352: In this case we require $f(E_\nu)$, the spectrum of three-body decay, which is
353: \begin{eqnarray}
354: f(E_\nu)&=&16 \gamma_\mu^5 \bigg( {3\over\gamma_\mu^5} - {4\over 3}(3+\beta_\mu^2)\zeta \bigg) \zeta^2 {1\over m_\mu}\,,\;{\rm for}\; 0 \le \zeta \le [1-\beta_\mu]/2\nonumber\\
355: &=&{ {5\over 3} + {4\over (1+\beta_\mu)^3} \bigg( {8 \zeta\over 3} - 3(1+\beta_\mu) \bigg) \zeta^2}{ \beta_\mu \gamma_\mu m_\mu }\,,\;{\rm for}\; [1-\beta_\mu]/2 \le \zeta \le [1+\beta_\mu]/2\;,\qquad
356: \end{eqnarray}
357: where $\zeta = ({E_\nu\over\gamma_\mu m_\mu})$ and $\beta_\mu = v_\mu/c$.
358: Since detection lies far above 70 MeV,
359: it suffices to approximate neutrinos
360: from pion decay as the same in number as those from subsequent muon decay.
361:
362: In the case of proton-photon interactions, we may use the expressions (2, 5, \&8) unaltered,
363: changing only the source function of pion production. This is now given by
364: (Bottcher \& Dermer 1999)
365: \begin{equation}
366: q_\pi (\gamma_\pi) = {c\over 2} \int_1^\infty d\gamma_p\, n_p(\gamma_p) \int^1_{-1} d\chi
367: \int_0^\infty d\epsilon\; n_\gamma(\epsilon) (1-\beta_p \chi) {d^2 \sigma\over d\gamma_\pi d \chi}\;,
368: \end{equation}
369: where $n_p$ and $n_{\gamma}$ are, respectively, the proton and photon distributions.
370: The result is a flat neutrino emissivity
371: $dN_\nu/dE\;dV\;dt$ until threshold, with a spectral index one higher than that of the proton
372: distribution afterwards, $s_\nu = s_p + 1$. Again, $\chi$ is the interaction angle cosine.
373:
374: The cross-section for charged photopion production (Mucke et al. 2000, and references
375: cited therein) features two prominent decay resonances (Fig. 1): the first, at $E_\Delta = 330$ MeV,
376: has a height of $2.45 \times 10^{-28}$ cm$^2$ and a decay width of 0.15 GeV. The second
377: resonance, at 710 MeV, has a decay width of 0.18 GeV---but, due to the steepness of the proton
378: distribution, it does not play a major role in the neutrino flux. Note that this height is
379: some factor of 1.5 greater than that assumed by Dermer and Bottcher (2003).
380: The cross-section for neutral photopion production shows only a single resonance
381: of height $3 \times 10^{-28}$ cm$^2$ and a decay width of 0.15 GeV. Our fitting function
382: for a resonance at mass M, of height $\sigma_0$ and of width $\Gamma$ is of Breit-Wigner form,
383: \begin{equation}
384: f(\epsilon^\prime) = {\sigma_0 \Gamma^2 \epsilon^\prime\over(\epsilon^\prime-M^2)^2 + \Gamma^2 \epsilon^\prime}\;,
385: \end{equation}
386: where $\epsilon^\prime = \gamma_p \epsilon (1-\mu)$ is the photon energy in the proton rest frame.
387: We have tested the calculation with a simple $\delta$-function resonance, however, and the
388: results are nearly identical.
389:
390: Finally, we require the synchrotron emission from injected electrons to set the level of
391: ambient relativistic protons. In practice this is evaluated numerically, but a sufficient
392: guide may be given by the approximation for the electron injection (Mannheim \& Schlickeiser 1994),
393: \begin{eqnarray}
394: q_e&=& {m_\pi\over 70\, m_e} q_{\pi^\pm} \bigg({E_\pi\over 70\;\hbox{MeV}}\bigg)
395: \sim {13\over 12}\sigma_{pp}\, c\, n_{H}\, n_{p0}(r) \bigg({m_p\over 24\, m_e} \bigg)^{s_{e0}-1}
396: (\gamma_e \beta_e)^{-s_{e0}}\;\; {\hbox{cm}}^{-3}\; {\hbox{s}}^{-1}\nonumber \\
397: \null&\equiv&K_{inj}\, {\gamma_e}^{-s_{e0}}\;.
398: \end{eqnarray}
399: Assuming the dominant loss mechanism is either inverse Compton scattering with a blackbody
400: photon background ($T_{CMB} = 2.73$ K), or radio synchrotron, radiative losses are given
401: by $-dE_e/dt = a_s E_e^2$, with the constant $a_s = (4/3) \sigma_T\, c\, n_e (\epsilon_{CMB}
402: + \epsilon_{B})/m_e c^2$. The energy density ($\epsilon_{CMB}$) in the CMB dominates
403: over that ($\epsilon_B$) in the magnetic field. The equilibrium distribution of
404: electrons due to injection against these losses is (Blumenthal and Gould 1970)
405: \begin{equation}
406: n(\gamma_e) = {K_{inj}\over m_e c^2 a_s (s_{e0}-1)}\; \gamma_e^{-(s_{e0}+1)} \;.
407: \end{equation}
408: The radio synchrotron emissivity (in units of energy per unit volume, per unit time,
409: per unit frequency) associated with this distribution is then
410: \begin{equation}
411: {dE\over dV\, d\nu\, dt} \approx 1.15\, \pi^2 {K_{inj}\, \alpha\, \hbar\,
412: \nu_B\over m_e c^2\, a_s (s_{e0}-1)} \bigg({\nu_B\over\nu} \bigg)^{s_{e0}/2}\;,
413: \end{equation}
414: where $\nu_B$ is the gyrofrequency. For a final handle on the electron
415: population in clusters based on the presence or absence of a hard X-ray excess,
416: the corresponding Compton scattering emissivity
417: off the CMB (in units of photon number per unit volume, per unit time, per unit energy) is
418: \begin{equation}
419: {dN_\gamma\over dV\, d\epsilon\,dt} = 1.8{r_0^2\over\hbar^3 c^2}{K_{inj}\over m_e c^2 a_s
420: (s_{e0}-1)} (kT_{CMB})^{(s_{e0}+6)/2} \epsilon^{-(s_{e0}+2)/2}\;,
421: \end{equation}
422: where $r_0$ is the classical electron radius, and we have used the fact that the hard
423: X-radiation is produced below the Klein-Nishina region to simplify the cross section.
424:
425: \subsection{Cluster modeling}
426: We now have spectra or distributions for each of the links in the decay cascade---electrons,
427: $\gamma$-rays, and neutrinos---and have derived expressions relating electron number to observed radio or nonthermal X-ray emission. It remains only
428: to assume a radial distribution for the background gases and for energetic protons
429: diffusing throughout the cluster. For simplicity we will adopt a single
430: point source of protons at the center of the cluster:
431: realistic morphology, time dependence, or non-Kolmogorov proton diffusion
432: through the cluster may seem to increase the accuracy of the model, but without a motivating
433: observation only serve to make it less reproducible. Below we describe in detail
434: how we calculate the $\gamma$-ray and neutrino fluxes from Coma, the results
435: of which are presented graphically in Fig. 2.
436:
437: We take the background gas within Coma to be described by a $\beta$ model
438: (Colless \& Dunn 1996) with $n_{core} = 3 \times 10^{-3}$ cm$^{-3}$,
439: $\beta = 0.7$, and $r_{core} = 0.25$ Mpc, for which
440: \begin{equation}
441: n(r) = n_c \bigg[ 1+ \bigg({r\over r_c} \bigg)^2 \bigg]^{-3\beta/2}\;.
442: \end{equation}
443: While the thermal and nonthermal electron populations are well determined
444: by X-ray and radio observations (Fig. 3), respectively, no similar tool directly
445: determines the nonthermal proton population. Here, we assume high-energy
446: proton injection takes place at a central source and these then diffuse
447: outward according to (Blasi \& Colafrancesco 1999)
448: \begin{equation}
449: {\partial n_p(E_p,r,t)\over\partial t} = D(E_p) \bigtriangledown^2 n_p(E_p,r,t) + Q(E_p) \delta(r) +
450: {\partial\over\partial E_p} b(E_p) n_p (E_p,r,t)\;.
451: \end{equation}
452: The highest possible energy attainable by protons is actually given (Dermer \& Berrington 2002)
453: by the duration of acceleration, rather than any loss term $b(E_p)$.
454: Loss, and an additional term representing proton escape from the cluster, we neglect.
455: Assuming the cluster is relaxed, we then may integrate to give
456: \begin{equation}
457: n_p(E_p,r) = {Q_p(E_p)\over D(E_p)}{1\over 4 \pi r} \;,
458: \end{equation}
459: where $Q_p(E_p) = Q_0 E_p^{-s_p}$, and $Q_0$ is normalized to synchrotron observations.
460:
461: The resonant diffusion of protons in magnetic field fluctuations with a power spectrum $P(k)$ in
462: wave number $k$ is given by (Blasi \& Colafrancesco 1999)
463: \begin{equation}
464: D(p) = 1/3 c r_{L}(p){B^2\over\int^\infty_{1/r_L(p)} P(k) dk}\;,
465: \end{equation}
466: where $r_L(p) = pc/eB$ is the Larmor radius. In this paper we take fluctuations of the
467: magnetic field to have a Kolmogorov spectrum,
468: \begin{equation}
469: P(k) = P(k_0) \bigg({k\over k_0} \bigg) ^{-5/3}\;,
470: \end{equation}
471: and assume the spectrum is normalized to
472: \begin{equation}
473: \int^\infty_{k_0} P(k) dk \approx B^2.
474: \end{equation}
475: Thus
476: \begin{equation}
477: P(k_0) = {2\over 3}{1\over k_0} B^2\;,
478: \end{equation}
479: and
480: \begin{equation}
481: D(p) = 1/3 c k_0^{-2/3} (eB)^{-1/3} E^{1/3}\;.
482: \end{equation}
483: A typical value for the smallest scale on which the magnetic field is homogeneous,
484: $d_0 = 1/k_0$, is $> 20$ kpc, since the magnetic field is being stirred by the virial
485: motion of galaxies within the cluster, for which $d_0 = (4 \pi/(3 N_{gal}))^{1/3}
486: R_{cluster}$.
487:
488: \section{Detection Counts}
489:
490: In our model, the number of photons above $E_\gamma$ detected with GLAST,
491: writing $L_{43}$ as the luminosity in units of $10^{43}$ ergs s$^{-1}$, is (Berrington \& Dermer 2005)
492: \begin{equation}
493: N(>E_\gamma) \sim 35\, t_{yr} (E_\gamma/\hbox{\textrm{GeV}})^{-1.04}\;{L_{43}\over 0.3}\;,
494: \end{equation}
495: where we use a cluster evolution time $t_{yr} = 0.7$ Gyr, and $L_{43} = 0.1$ through $0.7$
496: coinciding with $s_p = 2.1$ through $2.5$. The detection significance
497: $\eta_\sigma$ for Coma, assuming it is a point-like source (i.e., the point spread function for a given
498: photon energy is larger than the $18^\prime$ angular extent of Coma) is (Berrington \& Dermer 2005)
499: \begin{equation}
500: \eta_\sigma \sim 5.4 \sqrt{t_{yr}} \bigg({L_{43}\over 0.3} \bigg) (E_\gamma/\textrm{GeV})^{0.1}\;.
501: \end{equation}
502: This yields a $\eta_\sigma = 5 \sigma$ detection significance.
503:
504: We may also easily construct an argument for why neutrinos \emph{may} be detectable.
505: Suppose the $\gamma$-ray flux detected from the $\gamma$-ray bright Abell clusters is around
506: \begin{equation}
507: \Phi_\gamma(1 \;\textrm{TeV}) \approx 10^{-11} \ \textrm{photons} \ \textrm{cm}^{-2} \ \textrm{s}^{-1}
508: \end{equation}
509: at a few TeV (see Fig 4). This implies a similar number of muon neutrinos. Following
510: Gaisser \& Stanev (1984), the flux of muons produced in deep-inelastic interactions
511: between neutrinos and ice is nearly
512: \begin{equation}
513: {d\Phi_\mu\over dE_\mu} \approx {N_{Av} \sigma\over \alpha} E_\nu \Phi_\nu \approx
514: 10^{-20} \ \textrm{GeV}^{-1} \ \textrm{cm}^{-2} \ \textrm{s}^{-1}\;,
515: \end{equation}
516: with the deep-inelastic cross section $\sigma$, muon energy loss rate $\alpha$, and
517: Avogadro number $N_{Av}$. The muons leave the interaction with about 1/3rd the incoming neutrino
518: energy, so that
519: \begin{equation}
520: \Phi_\mu \approx \bar{E_\mu} {d\Phi_\mu\over dE_\mu} \approx 3 \times 10^{-18} \;\textrm{cm}^{-2} \ \textrm{s}^{-1}.
521: \end{equation}
522: The area of ICECUBE being $A\approx10^6$ m$^2$, this gives $N_{Ice} = 1
523: (t_{\rm obs}/{\rm year})$.
524: That is, the number of neutrino detection events per year is approximately
525: $\Phi_{\gamma}(1\;{\rm TeV})/10^{11}$
526: (see, e.g., Crocker, Melia, and Volkas 2000, 2002, 2005; Crocker, Fatuzzo, Jokipii et al.
527: 2005).
528: %This is borne out in a previous prediction (Crocker et al 2005) of neutrino flux based on HESS
529: %detection of the galactic center.
530: %*** I'm not clear on why this has been commented out. All I'm trying to say is that the neutrino
531: %*** number prediction should be something like $N_\gamma(GeV)/10^{-11}$ year$^{-1}$. Since
532: %*** even the interested reader is unlikely to go beyond the back-of-the-envelope I'd like
533: %*** to present at least one reference confirming this is the case.
534: %*** And, comparing the 2005
535: %*** paper's neutrino rate to the gamma flux levels, in looks as though in the case of the galactic center
536: %*** we've got that relation \emph{precisely}.
537:
538: More precisely, the number of neutrino detection events in a particular device is
539: \begin{equation}
540: N_{\rm year} = \int dE_\nu \int^{year} \ dt \ {\rm Area}[E_\nu,\theta(t)] \ \Phi(E_\nu) \times P_{\rm detect}(E_\nu) \ {\rm Attn}[E_\nu,\theta(t)]\;,
541: \end{equation}
542: which includes attenuation of the flux as it passes through Earth,
543: the effective muon area---the preceding
544: depending on the location of the source defined in terms of its angle from
545: the nadir $\theta$---and
546: the probability of neutrino interaction and subsequent detection.
547: For the ICECUBE detector, these quantities are specified in detail in the Preliminary Design Document
548: available from the ICECUBE website (http://icecube.wisc.edu/). See also
549: Ahrens et al. (2004).
550:
551: Carrying this out gives an event rate of $0.3$ detection events per year from Coma via the
552: ICECUBE detector. More optimistic are the detection rates should the $\gamma$-ray bright
553: Abell clusters be confirmed: these range between $30$ upcoming events per year from Abell
554: $1914$ in the most optimistic case of a $s_p=2.1$ spectral index, to $8\times10^{-2}$
555: events per year from an extremely strong shock in Abell 85 (for the most pessimistic case).
556:
557: Meanwhile, the prospect for using Auger to directly probe neutrino production via
558: p$\gamma$ collision processes in clusters is possible, but it would require very
559: favorable system parameters: even assuming
560: the weakest possible shock (and the most optimistic duty cycle of $15\%$),
561: yields only a few detection events per year, with Abell 1914 representing the
562: most promising source.
563:
564: Table 2, summarizing these results, as calculated on the basis
565: of the results presented by Miele et al. (2006),
566: is given below; here, the fields separated by a semicolon indicate
567: decreasing proton spectra, from $s_p = 2.1$ to $2.5$, in steps of $0.1$.
568: At the $\sim$EeV energy scale of the p$\gamma$ neutrinos,
569: other experiments may offer apertures superior to that of Auger.
570: In principle, the ANITA experiment (Barwick et al. 2006), in particular, has
571: a considerably larger neutrino aperture than Auger but, unfortunately,
572: restrictions on the positions of potential sources (which must be located
573: within a narrow band of declination $[-9^\circ,19^\circ]$ to be detectable)
574: exclude all the potential cluster neutrino source considered here.
575:
576:
577: \section{Detections from Known Radio Clusters}
578:
579: Let us suppose that $\gamma$-ray detections from Abell clusters 85, 1914, and 1758
580: are due to $\pi^0$ decay. We may then use multiwavelength observations to set
581: a very solid prediction for their neutrino flux, depending only on the spectral
582: index of protons---and their maximum energy. These are shown in Fig. 4 for a
583: range of proton spectral indices consistent with Fermi acceleration and the
584: synchrotron observations.
585:
586: The spectral slope of radio emission for Abell clusters 85, 1914, and 1758 are $>1.5$,
587: 1.19, and 1.13, respectively (Ensslin et al. 1998).
588: Abell 85's value of $\alpha = 1.5$ would require a proton spectral index above that which could be produced
589: by shock acceleration. This could indicate either variability in measurements made years apart, or
590: the influence of several bright radio galaxies within the cluster. Here we take the maximum
591: value reasonable for shock acceleration, $s_p = 2.5$. The other values indicate proton
592: injection spectra near $s_p = 2.3$, matching the cooled spectrum of Coma (see Eqs. 13-15).
593:
594: Of the three, Abell 85 appears to be the only cluster (Durret et al. 2005) for
595: which temperature maps exist to create a radial gas profile, although Chandra
596: detections of 1914 do exist (Govoni 2005). For Abell 85 the core gas density is
597: $7.7 \times 10^{-3}$ cm$^{-3}$, while in the other two clusters we assume it is
598: the same as in Coma ($n_e = 3 \times 10^{-3}$ cm$^{-3}$). The flux of particles
599: for each cluster scales as its relative redshift to Coma, while the active volume
600: scales as the largest linear scale (LLS). In every case we take a nominal $B=1 \mu$G,
601: for lack of better knowledge. This results in a strikingly similar value for the
602: CR proton number density required to fit the radio flux (Fig. 3): in every case
603: (including Coma) it is within a factor of two of $n_{CR} = 10^{-8}$ cm$^{-3}$, where
604: $n(p) = n_{CR} p^{-s_p}$. This is very near energy equipartition with the background
605: thermal electrons. This balance in energy between the two species may be just
606: a coincidence, but more likely it represents a dynamic coupling between the protons
607: and the ambient electrons, such that an effective energy transfer occurs between
608: the two (see Wolfe and Melia 2008).
609:
610: In each case,
611: $\gamma$-rays from $\pi^0$-decay dominate those from nonthermal bremsstrahlung by the synchrotron-emitting
612: electrons. It is therefore impossible to describe both the Coma cluster and these $\gamma$-ray bright clusters
613: within a simple secondary model, since Coma should then be a copious source of $\gamma$-rays. The situation
614: is not improved if one assumes nonthermal bremsstrahlung from the electrons produces the $\gamma$-rays,
615: since this would (via inverse Compton scattering) produce an X-ray excess not observed in Abell 1785 and 1914
616: (Abell 85, however, displays such an excess).
617:
618: De Marco et al. (2006) have advanced the notion that photopion production between cosmic-ray protons
619: and infrared light emitted by galaxies within the cluster may dominate over proton-CMB interactions.
620: We agree with this general result, but find some difference in the details.
621: Again, pion production from p$\gamma$ interactions is characterized by a hard cut-off at the
622: threshold energy, followed by a spectral index one greater than the proton index, $s_\pi = s_p+1$.
623: The three-body decay then gives an extra degree of freedom, allowing neutrinos to be produced at
624: a constant emissivity until threshold. The De Marco et al. (2006) (their Fig. 5) and our own
625: calculation include all of these features, except the location of threshold.
626: The energy of the galactic IR emission
627: under consideration (Lagache et al. 2003) is peaked a decade above the CMB peak; with the
628: threshold for $\gamma + p \rightarrow p + \pi$ being
629: \begin{equation}
630: E_{thres} = { (m_p + m_{\pi})^2 -m_p^2\over 4E_\gamma}\;,
631: \end{equation}
632: the Wein peak at $1.2 \times 10^{-3}$ eV gives a CMB threshold of $5.8 \times 10^{19}$ eV, while
633: the IR peak at $1.03 \times 10^{-2}$ eV gives a threshold of $6.6 \times 10^{18}$ eV (later
634: line emissions are suppressed by the steepness of the proton distribution). Yet the threshold
635: of the De Marco et al. (2006) calculation is $\sim 10^{16}$ eV. This shift in the emission peak
636: means neutrino production begins at a point in the proton distribution some 100 eV in energy above
637: that given in De Marco et al. (2006), which with $n(p) \sim E_p^{-s_p}$ corresponds to a factor of
638: $10^{-4}$ for the neutrino flux.
639:
640: Finally, we comment on the possibility of calculating \emph{diffuse} cluster $\gamma$-ray and neutrino fluxes,
641: using the Coma cluster as a template. In this case it is helpful to look at those radio halo/relic
642: clusters which are known to exist, assuming that bright sources of $\gamma$-rays or neutrinos will
643: have first been observed in one of the radio surveys.
644: If cosmic-ray production is dominated by cluster mergers, the number of CR protons,
645: and thus the radio luminosity (as well as neutrino and $\gamma$-ray production),
646: should roughly scale as a cluster's mass, i.e., $L_p \sim GM^2/Rt_{\rm merge}
647: \sim (M/M_{\rm Coma})^{5/3}$ (De Marco et al. 2006). As a given cluster's mass is virially
648: connected to its temperature, we may predict the flux of cosmic rays and
649: neutrinos as a function of that cluster's relative temperature and
650: distance with respect to Coma.
651: In this case, $M \sim 5T R_v/G (k/\mu m_p)$, where we take the virial radius as scaling
652: with the largest linear size of a cluster. The $M-T$ relation has also been derived
653: from specific Chandra observations (Kotov \& Vikhlinin 2006) as
654: $M = A (T/5)^\beta \times 10^{14} M_\odot$ with
655: $A = 1.7$ and $\beta = 1.5$.
656:
657: Should the population of radio halo/relics be powered by gravitational infall at the cluster's current mass,
658: we expect radio power to scale with the $5/3$rds power of cluster mass---but it does not
659: (Fig. 5). Using the Chandra $M-T$ relation gives a correlation between $(M/M_{\rm Coma})^{5/3}$
660: and $L_{\rm radio}$ of 0.46; the virial value improves this to 0.61. In no case is the relatively larger
661: proton density of a particular cluster greater than $(z_{\rm Coma}/z)^2$, meaning that none of
662: the known radio clusters could contribute a neutrino or $\gamma$-ray flux above that of Coma's.
663: The most concrete estimation for the diffuse flux---the weighted sum (by radio luminosity)
664: of the 30 or so known radio clusters---is approximately $9/4$ths that of Coma alone.
665: Of these, $\gamma$-ray bright double clusters Abell 1758 and Abell 1914 (Fig. 5, square
666: and diamond, respectively) are among the most dominant. The entire discussion
667: is moot if the identification of Abell 85, 1758, and 1914 as $\gamma$-ray
668: bright clusters is confirmed, in which case these three clusters alone dominate the diffuse flux
669: despite their relatively low mass; in this case the importance of ongoing, violent mergers is emphasized.
670: A generalization which would allow a solid diffuse flux to be calculated appears to depend on a present
671: lack of knowledge about the basic mechanisms of cluster halos.
672:
673: \section{Conclusions}
674:
675: Observations of both $\gamma$-rays ($\gamma_{pp}$) and neutrinos (both $\nu_{pp}$ and $\nu_{p\gamma}$)
676: in clusters provide answers to a series of yes-or-no questions. If both
677: $\gamma_{pp}$ and $\nu_{pp}$ are observed from a cluster, then proton-proton induced
678: cascades are certainly at work and their contribution to the radio halo can be evaluated---as well
679: as potential models for the creation of viable proton populations with total energies $>10^{61}$ erg.
680: If $\gamma_{pp}$ is detectable by either EGRET or GLAST and $\nu_{pp}$ is not observed,
681: then $\gamma_{pp}$ is not likely to be caused by secondary decay (this is not a firm `no' as the
682: source in question may occupy an unfavorable position in the night sky, or the proton population
683: may experience a coincidental cutoff just after 100 MeV) and new models would need to be considered.
684: An Auger $\nu_{p\gamma}$ detection would confirm our expectation that very energetic cosmic-ray
685: photons interact with the CMB to produce debris particles, thereby losing their energy.
686: This would add some support to the view that cosmic ray events above $6 \times 10^{19}$ eV
687: cannot be hadronic if the incident particles originate from beyond the GZK limit.
688: Finally, if $\nu_{pp}$ is observed without $\nu_{p\gamma}$, we may assume that the spectral index of protons
689: is steeper than $2.1$, or that the cluster is incapable of accelerating protons to this energy.
690:
691: The combination of X-ray, radio, $\gamma$-ray, and neutrino observations on galaxy clusters
692: would give a vital new perspective on the mechanisms producing high-energy particles in clusters.
693: A simultaneous detection of $\gamma$-rays and neutrinos from pp scatterings in a cluster would
694: confirm conclusively that secondary production is at work, and provide a new probe for
695: cosmic-ray production in the MeV and TeV ranges, as well as shock acceleration in clusters
696: and astrophysically in general. A detection (or upper limit) of p$\gamma$ neutrinos,
697: meanwhile, would demonstrate whether clusters are capable of accelerating protons to ultra-high
698: energies, and, indeed, how particle physics tested in colliders evolves towards these energies.
699:
700: A confirmation via GLAST of Abell 85, 1758, or 1914, as $\gamma$-ray bright
701: would confirm them as strong neutrino sources, potentially detectable in
702: the ICECUBE and, perhaps, Auger experiments. If a nonthermal proton population
703: approaching energy equipartition is demonstrated, these then form a dominant
704: source of pressure in clusters, affecting structure and evolution. Of course,
705: the fact that neutrino detections are only possible near equipartition means
706: that there may be only a narrow window of opportunity for actually measuring
707: a neutrino flux, or maybe none at all. But we find it promising that an
708: interesting prediction may be made even without requiring a super-equipartition
709: proton population. And, anyway, a near-equipartition situation between the
710: protons and the background electrons is suggested by a dynamic coupling between
711: these two populations, as discussed in Wolfe and Melia (2008). Taken together
712: with evidence of ongoing mergers, these observations would give new insight
713: into the period of cluster formation, beginning some $>9$ billion years ago,
714: and what role mergers play in cosmic-ray production.
715:
716: \acknowledgements
717:
718: This research was supported by NSF grant AST-0402502 at the University of Arizona. This
719: work was supported in part by the Australian Research Council.
720: BW and FM are grateful for the hospitality of the Universities of Melbourne and Canterbury,
721: where a portion of this work was carried out.
722:
723:
724: \begin{thebibliography}{}
725: \bibitem[Ahrens et al. 2004]{one} Ahrens et al., 2004 AstroParticle Phys., 20, 507.
726: \bibitem[Ando \& Nagai 2008]{one.a} Ando, S., and Nagai, D., 2008, MNRAS 385, 2243
727: \bibitem[Berrington and Dermer 2005]{two} Berrington, R. C. and Dermer, C. C., 2003, ApJ,
728: 594, 709
729: \bibitem[Barwick et al. 2006]{three} Barwick et al., 2006, PRL 96, 171101
730: \bibitem[Berezinsky et al. 1997]{three.a} Berezinsky, V. S., Blasi, P., and
731: Ptuskin, V. S., 1997, ApJ, 487, 529
732: \bibitem[Berrington \& Dermer 2003]{three.b} Berrington, R. C., and Dermer,
733: C. D., 2003, ApJ, 594, 709
734: \bibitem[Blasi 2004]{three.c} Blasi, P., 2004, JKAS, 37, 483
735: \bibitem[Blasi and Colafrancesco 1998]{three.d} Blasi, P. and Colafrancesco, S.,
736: 1998, Astropart. Phys., 9, 227
737: \bibitem[Blasi and Colafrancesco 1999]{four} Blasi, P. and Colafrancesco, S., 1999, Astropart. Phys.,12, 169
738: \bibitem[Blasi and Colafrencesco 2000]{four.a} Blasi, P., and Colafrancesco, S.,
739: 2000, Nuclear Physics B Proceedings Supplements, 80, 8
740: \bibitem[Blasi et al. 2007]{four.b} Blasi, P., Gabici, S., and Brunetti, G., 2007,
741: Int.J.Mod.Phys., A22, 681
742: \bibitem[Blumenthal \& Gould 1970]{five} Blumenthal, G. and Gould, R., 1970, Rev. Mod. Phys. 42, 237
743: \bibitem[Bottcher and Dermer 1998]{six} Boettcher, M. and Dermer, C., 1998, ApJ, 499, 131
744: \bibitem[Brunetti et al. 2001]{seven} Brunetti, G., Setti, G., Feretti, L., and
745: Giovannini, G., 2001, MNRAS, 320, 365
746: \bibitem[Brunetti et al. 2004]{eight} Brunetti, G., Blasi, P., Cassano, R., and Gabici, S., 2004, MNRAS, 350, 1174
747: \bibitem[Brunetti and Blasi 2005]{eight.a} Brunetti, G., and Blasi, P., 2005,
748: MNRAS, 363, 1173
749: \bibitem[Brunetti et al. 2007]{eight.b} Brunetti, G., Venturi, T., Dallacasa, D.,
750: Cassano, R., Dolag, K., Giacintucci, S., and Setti, G., 2007, ApJ Lett., 670, L5
751: \bibitem[Colafrancesco 2002]{nine} Colafrancesco, S., 2002, A\&A, 396, 31
752: \bibitem[Colafrancesco et al. 2004]{nine.a} Colafrancesco, S., Dar, A., and De R\'ujula, A.,
753: 2004, A\&A, 413, 441
754: \bibitem[Colless \& Dunn 1996]{ten} Colless, M. and Dunn, A., 1996, ApJ, 458, 435
755: \bibitem[Crocker, Fatuzzo, Jokipii et al. 2005]{eleven.b} Crocker, R., Fatuzzo, M., Jokipii, R., Melia,
756: F., and Volkas, R., 2005, ApJ, 622, 892
757: \bibitem[Crocker, Melia, and Volkas 2000]{eleven} Crocker, R., Melia, F., and Volkas, R., 2000, ApJS, 130, 339
758: \bibitem[Crocker, Melia, and Volkas 2002]{eleven.a} Crocker, R., Melia, F., and Volkas, R., 2002, ApJS, 141, 147
759: \bibitem[Crocker, Melia, and Volkas 2005]{eleven.c} Crocker, R., Melia, F., and Volkas, R., 2005,
760: ApJ Lett., 622, L37
761: \bibitem[David \& Kempner 2004]{twelve} David, L. and Kempner, J., 2004, ApJ, 613, 831
762: \bibitem[De Marco et al 2006]{thirteen} De Marco, D. et al., 2006, Phys. Rev. D, 73, 3004
763: \bibitem[Dennison 1980]{fourteen} Dennison, B., 1980, ApJ, 239, L93
764: \bibitem[Dermer 1986]{fifteen} Dermer, C., 1986, ApJ, 307, 47
765: \bibitem[Dermer and Barrington 2003]{sixteen} Berrington, R. C., Dermer, C. D., 2003, ApJ, 594, 709
766: \bibitem[Dolag and Ensslin 2000]{sixteen.a} Dolag, K., and Ensslin, T. A., 2000, A\&A, 362, 151
767: \bibitem[Durret et al 2005]{seventeen} Durret, F., Lima Neto, G., and Forman, W., 2005, A\&A, 432, 809
768: \bibitem[Ebeling et al. 1998]{eighteen} Ebeling, H. et al, 1998, MNRAS, 301, 881
769: \bibitem[Ensslin et al. 1998]{eighteen.a} Ensslin, T. et al., 1998, A\&A, 333, 47
770: \bibitem[Ensslin et al. 2007]{eighteen.b} Ensslin, T. A., Pfrommer, C., Springel, V.,
771: and Jubelgas, M., 2007, A\&A, 473, 41
772: \bibitem[Fatuzzo \& Melia 2003]{nineteen} Fatuzzo, M. and Melia, F., 2003, ApJ, 596, 1035
773: \bibitem[Gaisser \& Stanev 1985]{twenty} Gaisser, T. and Stanev, T., Phys. Rev. Lett., 1985, 54, 2265
774: \bibitem[Giovannini et al. 1999]{twentyone} Giovannini, G, Tordi, M., and Feretti, L., 1999, New Astr., 4, 141
775: \bibitem[Govoni et al. 2004]{twentytwo} Govoni, F. et al., 2004, ApJ, 609, 617
776: \bibitem[Guo \& Oh 2008]{twentytwo.a} Guo, F., and Oh, S. P., 2008, MNRAS, 384, 251
777: \bibitem[Hinton 2006]{twentytwo.b} Hinton, J. and Domainko, W., 2006, MNRAS, submitted (astro-ph/0701033v2)
778: \bibitem[Iyudin et al. 2004]{twentytwo.c} Iyudin, A. F., B\"ohringer, H., Dogiel, V.,
779: and Morfill, G., 2004, A\&A, 413, 817
780: \bibitem[Jones 2004]{twentytwo.d} Jones, T. W., 2003, JKAS, 37, 421
781: \bibitem[Jubelgas et al. 2008]{twentytwo.e} Jubelgas, M., Springel, V., Ensslin, T.,
782: and Pfrommer, C., 2008, A\&A 481, 33
783: \bibitem[Kempner \& Sarazin 2001]{twentythree} Kempner, J. and Sarazin, C., 2001, 548, 639
784: \bibitem[Kuo et al. 2005]{twentythree.a} Kuo, P.-H., Bowyer, S., and Hwang, C.-Y., 2005,
785: ApJ, 618, 675
786: \bibitem[Lima-Neto 2001]{twentythree.b} Lima-Neto, G., Pislar, V., and Bagchi, J., 2001, A\&A, 368, 440
787: \bibitem[Liu, Melia, and Petrosian 2006]{twentythree.c} Liu, S., Melia, F., and Petrosian, V.,
788: 2006, ApJ, 636, 798
789: \bibitem[Liu, Petrosian, and Melia 2004]{twentythree.d} Liu, S., Petrosian, V., and Melia, F.,
790: 2004, ApJ Lett., 611, L101
791: \bibitem[Mannheim \& Schleickeiser 1994]{twentyfour} Mannheim, K. and Schlickeiser, R., 1994, A\&A, 286, 983
792: \bibitem[Marchegiani et al. 2007]{twentyfour.a} Marchegiani, P., Perola, G. C.,
793: and Colafrancesco, S., 2007, A\&A, 465, 41
794: \bibitem[Marscher et al 1980]{twentyfive} Marscher, A., Vestrand, W., and Scott, J., 1980, ApJ, 241,1166
795: \bibitem[Markoff 1997]{twentyfive.a} Markoff, S., Melia, F., and Sarcevic, I., 1997, ApJ Lett., 489, L47
796: \bibitem[Markoff 1999]{twentysix} Markoff, S., Melia, F., and Sarcevic, I., 1999, ApJ, 522, 870
797: \bibitem[McNamara 2005]{twentyseven} McNamara, B., Wise, B., and Nulsen, P., 2005, Nature, 433, 45
798: \bibitem[Miele et al. 2006]{twentyeight} Miele et al., 2006, Phys. Lett. B, 634, 137.
799: \bibitem[Miniati et al. 2001]{twentyeight.a} Miniati, F., Ryu, D., Kang, H., and
800: Jones, T. W., 2001, ApJ 559, 59
801: \bibitem[Moffeit 1972]{twentynine} Moffeit et al., 1972, Phys. Rev. D., 5, 1603
802: \bibitem[Mucke et al. 2000]{thirty} Mucke, A. et al., 2000, Comput. Phys. Commun., 124, 290
803: \bibitem[Nakar et al. 2008]{thirty.a} Nakar, E., Milosavljevi\'c, M., and Nagai, D.,
804: 2008, ApJ, 675, 126
805: \bibitem[Neumann 2001]{thirtyone} Neumann, D. et al., 2001, A\&A, 365, L74
806: \bibitem[Nulsen 2005a]{thirtytwo} Nulsen, P., Hambrick,D., and McNamara, B., 2005a, ApJ, 625, 9
807: \bibitem[Nulsen 2005b]{thirtythree} Nulsen, P., McNamara, B., and Wise, B., 2005b, ApJ, 628, 629
808: \bibitem[Petrosian \& Bykov 2008]{thirtythree.a} Petrosian, V., and Bykov, A. M., 2008,
809: Space Science Reviews, 134, 207
810: \bibitem[Pfrommer \& Ensslin 2004]{thirtythree.b} Pfrommer, C., and Ensslin, T. A., 2004, A\&A, 413, 17
811: \bibitem[Reimer et al. 2003]{thirtyfour} Reimer, O. et al., 2003, ApJ, 588, 155
812: \bibitem[Reimer et al. 2004]{thirtyfour.a} Reimer, A., Reimer, O., Schlickeiser, R.,
813: and Iyudin, A., 2004, A\&A, 424, 773
814: \bibitem[Sarazin 2004]{thirtyfive} Sarazin, C. L., 2004, J. Kor. Astr. Soc., 37, 433
815: \bibitem[Sarazin 2007]{thirtyfive.a} Sarazin, C. L., 2007, Highlights of Astronomy, 14, 369
816: \bibitem[Stecker 1979]{thirtysix} Stecker, F. W., 1971, Cosmic Gamma Rays (Baltimore:Mono)
817: \bibitem[Wolfe and Melia 2006]{thirtysix.b} Wolfe, B. and Melia, F., 2006, ApJ, 638, 125 (arXiv:astro-ph/0505041)
818: \bibitem[Wolfe and Melia 2008]{thirtysix.c} Wolfe, B. and Melia, F., 2008, ApJ, 675, 156 (arXiv:0712.0187)
819: \bibitem[Zatsepin 1962]{thirtyseven} Zatsepin, G. and Kuz'min, V., 1962, Sov. Phys. JETP, 14, 1294
820: \end{thebibliography}
821:
822: \newpage
823: %\begin{landscape}
824: \begin{table}[h1]
825: \caption{Properties of Clusters Containing Halo and/or Relic Candidates\label{tbl-1}}
826: \scriptsize
827: \begin{tabular}{lccccccccccc}
828: \tableline\tableline\\
829: Name & z & RA & DEC & L$_X$ (0.1-2.4 keV) & T & S$_{327}$ & S$_{1400}$ & Radio Power & LLS & L$_\gamma$
830: \tabularnewline
831: & & (h m s) & ($^o$ $\prime$ $\prime\prime$) & ($10^{44}$ erg s$^{-1}$) & (keV)
832: & (mJy) & (mJy) & $10^{24}$ W Hz$^{-1}$ & kpc &$10^{-8}$ cm$^{-2}$ s$^{-1}$\\
833: \tableline
834: A665 &0.1818 &08 30 47.4 &65 51 14 &14.78 &8.3 &108 &16 &17.6 &1900\\
835: A697 &0.282 &08 42 57.6 &36 21 59 &16.30 &10.5 &29 &7 &11.2 &920\\
836: A725 &0.0921 &09 01 10.1 &62 37 20 &0.80 &7.3 &76 &6 &3.1 &440\\
837: A773 &0.2170 &09 17 54.0 &51 42 58 &12.35 &9.2 &35 &8 &7.9 &1400\\
838: A786 &0.1241 &09 28 49.7 &74 47 55 &1.53 &10.8 &319 &104 &21.9 &1400\\
839: A796 &0.1475 &09 28 00.0 &60 23 00 &1.38 &6.3 &53 &8 &5.6 &1200\\
840: A1758 &0.2800 &13 32 45.3 &50 32 53 &11.2 &7.2 &55 &11 &3.6 &1300 &9.2\\
841: A1914 &0.1712 &14 26 02.2 &37 50 06 &17.93 &10.7 &114 &20 &6.2 &1500 &16.3\\
842: A2034 &0.1130 &15 10 11.7 &33 29 12 &6.86 &7.0 &44 &8 &1.0 &920\\
843: A2061 &0.0777 &15 21 17.0 &30 38 24 &3.92 &5.5 &104 &19 &21.5 &920\\
844: A2218 &0.1710 &16 35 52.8 &66 12 59 &8.77 &6.7 &9 &1 &16.1 &340\\
845: A2219 &0.2281 &16 40 22.5 &46 42 22 &19.80 &11.2 &19 &2 &2.6 &810\\
846: A2255 &0.0809 &17 12 45.1 &64 03 43 &5.68 &7.3 &360 &18 &2.9 &930\\
847: A2256 &0.0581 &17 04 02.4 &78 37 55 &6.99 &7.3 &1165 &190 &1.3 &1450\\
848: A2319 &0.0555 &19 21 05.8 &43 57 50 &12.99 &9.9 &204 &32 &5.3 &580\\
849: A1656C &0.0232 &12 57 26.6 &28 15 16 &7.26 &8.2 &3081 &640 &11.5 &1500\\
850: A1367 &0.0215 &11 40 18 &20 18 01 &1.6 &3.5 & &182 \\
851: A13 &0.0943 &00 13 32.2 &219 30 03.6 &2.24 &4.3 &&34 &1.3 &880\\
852: A2744 &0.3080 &00 14 16.1 &230 22 58.8 &22.05 &11.04 &&38 &15.5 &1700\\
853: A85 &0.0555 &00 41 48.7 &209 19 04.8 &8.38 &6.2 &&46 &0.61 &480 &12\\
854: A115 &0.1971 &00 55 59.8 &126 22 40.8 &14.57 &9.8 &&80 &1.3 &1500\\
855: A401 &0.0739 &02 58 56.9 &113 34 22.8 &9.88 &7.8 &&25 &0.6 &590\\
856: A520 &0.2030 &04 54 07.4 &102 55 12.0 &14.20 &8.33 &&38 &6.7 &1080\\
857: A545 &0.1540 &05 32 23.3 &211 32 09.6 &9.29 &5.5 &&41 &4.4 &1500\\
858: A548 &0.0424 &05 45 27.8 &225 54 21.6 &0.30 &2.4 &&50 &0.4 &360\\
859: A1300 &0.3071 &11 31 54.9 &219 54 50.4 &23.40 &5 &&14 &5.7 &780\\
860: A1664 &0.1276 &13 03 44.2 &224 15 21.6 &5.36 &6.5 &&107 &7.5 &1400\\
861: A2163 &0.2080 &16 15 49.4 &206 09 00 &37.50 &13.83 &&55 &10.2 &1500\\
862: A2254 &0.1780 &17 17 46.8 &119 40 48.0 &7.19 &7.2 &&32 &4.4 &1140\\
863: A2345 &0.1760 &21 26 58.6 &212 08 27.6 &9.93 &8.2 &&92 &12.3 &1200\\
864: A2390 &0.2329 &21 53 36.7 &117 41 32.2 &21.25 &10.13 &&69 &9.2 &1560\\
865: \tableline\tableline\\
866: \end{tabular}
867: \end{table}
868: %\end{landscape}
869:
870: \clearpage
871:
872: \begin{table}[h2]
873: \normalsize
874: \caption{Neutrino Detection Event Rates\label{tbl-2}}
875: \begin{tabular}{lccc}
876: \tableline\tableline\\
877: Name & IC upcoming (1/year) & IC downgoing (1/year) & Auger (1/decade)\\
878: \tableline
879: Abell 85 & -- & 0.08; 0.3; 1; 3; 13 & 7e-5; 9e-4; 1e-2; 2e-1; 2\\
880: Abell 1914 & 0.2; 0.8; 2; 8; 30 & -- & 9e-5; 1e-3; 2e-3; 2e-1; 3\\
881: Abell 1758 & 0.1; 0.4; 1; 5; 18 & -- & 5e-5; 7e-4; 9e-3; 1e-1; 2\\
882: Coma & 0.3 & -- & 0.03\\
883: \end{tabular}
884: \end{table}
885:
886: \clearpage
887:
888: \begin{figure}
889: \centering
890: \plotone{f1.eps}
891: \caption{
892: Cross-section for charged pion production via p$\gamma$ scattering, and its fit.
893: \emph{dashed}--the fit for the $\Delta$ resonance; \emph{dotted}--fit to the
894: $\epsilon^\prime \sim 0.7$ GeV resonance; \emph{solid}--overall fit;
895: \emph{stars} data from Mucke et al. (2000).
896: }
897: \end{figure}
898:
899: %\begin{figure}
900: %\centering
901: %\plotone{f2.eps}
902: %\caption{
903: %Gamma-ray emission from Coma (thin solid) depends on the proton spectral index $s_p$, here displayed
904: %between $s_p = 2.1$ through $2.5$. It also depends on the radio emission, since a
905: %change in $s_p$ means renormalizing the CR proton density to fit these observations (thus
906: %$\gamma$ ray emission in the 70 MeV region increases with increasing spectral index).
907: %Also pictured are the detection limits for GLAST (1 year, dashed), EGRET (1 year, dash-dotted),
908: %VERITAS (thick solid), and MAGIC (5 $\sigma$, 50 hrs, dotted); as well as the nonthermal bremsstrahlung
909: %emission from the synchrotron-emitting electrons at $s_p = 2.1$ (solid shaded). $\pi^0$ decays will
910: %dominate $\gamma$ ray emission unless $B\ll 1 \mu$G.
911: %}
912: %\end{figure}
913:
914: \clearpage
915:
916: \begin{figure}
917: \centering
918: \plotone{f2.eps}
919: \caption{
920: Gamma rays (left panel) and associated muon neutrinos (right panel) from the Coma cluster. In the left panel,
921: a flux of $L_{\gamma, 43} = 0.3$ is given by pion decay (thin solid),
922: which wholly dominates nonthermal bremsstrahlung from synchrotron-emitting electrons (solid shaded).
923: Also pictured are the detection limits for GLAST (1 year, dashed), EGRET (1 year, dash-dotted),
924: VERITAS (thick solid), and MAGIC (5 $\sigma$, 50 hrs, dotted). In the right panel we show Coma's associated
925: neutrino flux for pp scattering (dash-dotted), p$\gamma$ scattering with the IRB (solid shaded)
926: and that with the CMB (thick solid). The $\gamma$-ray detection limits (dotted) and flux (thin solid)
927: are shown for comparison.
928: }
929: \end{figure}
930:
931: \clearpage
932:
933: \begin{figure}
934: \centering
935: \plotone{f3.eps}
936: \caption{
937: Radio emission from the four clusters sets the scale for the number and
938: index of cosmic rays.
939: \emph{solid}--Coma;
940: \emph{dashed}--Abell 1914;
941: \emph{dotted}--Abell 1758;
942: \emph{dash-dot}--Abell 85.
943: In each of the four clusters examined here, the number of cosmic-ray
944: protons is a few times $10^{-8}$ cm$^{-3}$, placing them just below energy equipartition
945: with the thermal background. The implied similarity of the three clusters' environments
946: (other than Coma)
947: means that, in each case, $\gamma$-ray emission is dominated by $\pi^0$ decay.
948: }
949: \end{figure}
950:
951: \clearpage
952:
953: \begin{figure}
954: \centering
955: \plotone{f4.eps}
956: \caption{
957: A positive $\sim$ GeV $\gamma$-ray detection directly implies a neutrino flux. In this case we display a
958: model for Abell 1914 which describes its $\gamma$-ray flux as a secondary model (solid, thin),
959: leaving the precise spectral index ($s_p= 2.1$ through $2.5$, in steps of $0.1$, moving downwards)
960: as an unknown. The EGRET and GLAST limits (dotted) are again reproduced as a guide.
961: Neutrinos produced in concert with $\gamma$-ray emission from pp scattering (exclusively
962: $\nu_\mu$ produced via muon decay, dash-dotted)
963: dominate to several decades of TeV; those produced in p$\gamma$ interactions
964: with an infrared background (solid, shaded) peak at a few times $10^{18}$ eV
965: and dominate at a few times PeV; and p$\gamma$ interactions with the CMB (thick solid)
966: dominate above EeV.
967: }
968: \end{figure}
969:
970: \clearpage
971:
972: \begin{figure}
973: \centering
974: \plotone{f5.eps}
975: \caption{
976: Radio luminosity of radio halo/relics does not scale as their mass. It is therefore difficult
977: to produce a template for calculating the total diffuse flux. The $\gamma$-ray bright radio halos
978: Abell 1758 (box) and 1914 (diamond) are relatively more massive than other clusters, which may
979: be why they are observed via EGRET.
980: }
981: \end{figure}
982:
983:
984:
985: \end{document}
986:
987: