1:
2: \documentclass[11pt,preprint]{aastex}
3:
4: %%\documentclass{emulateapj}
5:
6: %\documentclass[preprint]{aastex}
7: %\documentclass[manuscript]{aastex}
8:
9: %\documentclass[preprint2]{aastex}
10:
11: %\usepackage{emulateapj}
12:
13: \usepackage{natbib,graphicx,epsfig}
14:
15:
16: \newcommand{\psim}{\lower.5ex\hbox{$\; \buildrel \propto \over\sim \;$}}
17: %\def\gtrsim{\lower.5ex\hbox{$\; \buildrel > \over\sim \;$}}
18: % \def\lesssim{\lower.5ex\hbox{$\; \buildrel < \over\sim \;$}}
19: \def\e{{\epsilon}}
20: \def\ag{\alpha_{\gamma}}
21: \def\g{\gamma}
22:
23: \newcommand{\pr}{^\prime}
24: \newcommand{\gr}{ $\gamma$-ray }
25: \newcommand{\el}{\ell_{\rm S}}
26:
27:
28: \shorttitle{Gamma Rays from UHECRs in Cygnus A}
29: \shortauthors{Atoyan \& Dermer}
30:
31: %\received{}
32:
33:
34: \begin{document}
35:
36: \title{Gamma Rays from Ultra-High Energy Cosmic Rays in Cygnus A}
37:
38: \author{Armen Atoyan\altaffilmark{1} and Charles D. Dermer\altaffilmark{2}}
39: \altaffiltext{1}{Department of Mathematics, Concordia University,
40: 1455 de Maisonneuve Blvd.\ West,\\ Montreal, Quebec H3G 1M8, Canada;
41: ~atoyan@mathstat.concordia.ca}
42:
43: \altaffiltext{2}{Space Science Division,
44: Code 7653, Naval Research Laboratory,\\ Washington, DC 20375-5352, USA;~
45: charles.dermer@nrl.navy.mil}
46:
47:
48: \begin{abstract}
49: Ultra-high energy cosmic rays (UHECRs) accelerated in the jets of active galactic
50: nuclei can accumulate in high magnetic field, $\sim 100$ kpc-scale regions
51: surrounding powerful radio galaxies. Photohadronic processes { involving
52: UHECRs and photons of the extragalactic background light}
53: make ultra-relativistic electrons and positrons that initiate
54: electromagnetic cascades,
55: leading to the production of a $\gamma$-ray synchrotron halo.
56: We calculate the halo emission in the case of Cygnus A and show that
57: it should be detectable with the { Fermi Gamma ray Space Telescope}
58: and possibly detectable with ground-based
59: $\gamma$-ray telescopes if radio galaxies are the sources of UHECRs.
60: \end{abstract}
61:
62: \keywords{Galaxies: individual (Cygnus A) --- cosmic rays: theory ---
63: galaxies: active --- galaxies: jets --- radiation processes: nonthermal}
64:
65: %\maketitle
66:
67: %\vspace{3mm}
68:
69: \section{Introduction}
70:
71: Active galactic nuclei (AGN) and gamma-ray bursts are considered
72: as two of the most plausible classes of astrophysical accelerators of
73: extragalactic ultra-high energy cosmic rays \citep[UHECRs; see, e.g.,][]{hh02}.
74: The recent report of the \citet{Auger07} about clustering of the arrival directions
75: of UHECRs with energies $E \gtrsim 6\times 10^{19}$ eV within $\approx 3^\circ$
76: of the directions to AGN at distances $d\lesssim 75\, \rm Mpc$
77: strongly suggests that effective production of cosmic rays
78: with energies $E\sim 10^{20}\,\rm eV$ takes place in at least one of these
79: source classes. Because of photohadronic GZK interactions of protons or ions
80: with the CMB radiation,
81: the study of super-GZK UHECRs from sources at $d\gtrsim 100\,\rm Mpc$ becomes impossible
82: with cosmic-ray detectors like the Pierre Auger Observatory \citep{hmr06}.
83: Powerful AGN, as well as GRBs, are mostly located at larger distances.
84:
85: { Relativistic beams of energy from the central nuclei of AGN are thought to
86: power the multi-kpc scale radio lobes of powerful galaxies and form
87: an extended cavity \citep{sch74}.
88: Acceleration of UHECRs in the compact inner jets of the radio galaxy
89: on the sub-parsec scale, followed by
90: production of collimated beams of ultra-high energy neutrons and
91: gamma-rays, provides a specific mechanism to transport energy to
92: the radio lobes and cavity \citep{ad03}.}
93: When the neutron-decay UHECR protons interact with the extragalactic background light (EBL),
94: which is dominated by the CMB radiation,
95: ultra-relativistic electrons (including positrons) and $\gamma$ rays with
96: $E\gtrsim 10^{18}\,\rm eV$ are produced. Such secondaries can initiate
97: pair-photon cascades to form large multi-Mpc scale halos of GeV/TeV radiation due
98: to Compton and synchrotron processes \citep{acv94,aha02,ias05}.
99: %Moreover, if UHECRs escape from the central galaxy directly into intergalactic space
100: %and travel toward the observer in weak intergalactic magnetic fields
101: %at the 1 -- 100 nG level,
102: %point-like GeV -- TeV synchrotron sources could be formed
103: %\citep{ga05} when UHECRs lose energy through photomeson
104: %processes with photons of the EBL.
105: %Production of $\gamma$-ray fluxes induced by UHECRs
106: %in the weak, $\lesssim 1~\mu$G magnetic fields of nearby clusters of galaxies, so that
107: %UHECR deflection could still be neglected, was also considered by \citet{asm06}.
108: % { though without predictions for specific clusters}.
109:
110:
111: In this Letter, we predict that synchrotron GeV fluxes from UHECR AGN sources are
112: detectable with the Fermi Gamma ray Space Telescope (FGST;
113: formerly the Gamma ray Large
114: Area Space Telescope, GLAST) if UHECRs are captured in the
115: vicinity of radio galaxies for sufficiently long times.
116: Magnetic fields at the
117: $\gtrsim \mu$G level in the
118: kpc -- Mpc vicinity from the AGN core are required to
119: isotropize UHECRs accelerated by jets of radio galaxies. Indeed, for protons with
120: energy $E\equiv 10^{20} E_{20} \,\rm eV$, the mean magnetic field $B$ required
121: to provide gyroradii smaller than size $r$ is
122: $B\gtrsim 10^{-4} E_{20} r_{\,\rm kpc}^{-1}\,\rm G$, where
123: $r_{\rm kpc} =r/1\,\rm kpc$ is the spatial scale where the isotropization occurs.
124:
125: Here we consider the specific case of the powerful radio galaxy Cygnus A,
126: { where the mean magnetic field in the surrounding cavity could reach 10 -- 100 $\mu$G
127: at 100 kpc scales}. Its properties are considered in Section 2, and calculations are presented in Section 3.
128: %In Section 2, we discuss the properties of Cygnus A and
129: %show that if UHECRs are accelerated
130: %in the jets of this radio galaxy, then they can be trapped by the
131: %strong magnetic
132: %field in the vicinity of this source. In Section 3, calculations of high-energy
133: %electromagnetic cascade radiation induced by photohadronic processes
134: %are presented.
135: We summarize in Section 4.
136:
137: \section{Model Parameters}
138:
139: Magnetic fields $B\gtrsim 1\, \mu G$ could be present at $\lesssim 1\,$Mpc scales
140: in clusters of galaxies \citep{kim90,fer95,fer08}. Even higher magnetic
141: fields, $B\sim 10^{-4}\,$G, could be present in the $r\lesssim 100\,$kpc
142: vicinity of cD galaxies near the center of galaxy clusters like the powerful radio
143: galaxy Cyg A \citep{wil06}. In such magnetic fields, synchrotron radiation of
144: electrons with $E\sim 10^{18} \,\rm eV$
145: produced in photomeson interactions is in the TeV domain. Furthermore,
146: synchrotron radiation of lower energy electrons produced in
147: $p+\gamma\rightarrow p+ e^+ + e^-$ interactions by UHECRs extends to the GeV domain and could become
148: detectable with the Fermi Telescope.
149: Here we consider the detectability of
150: these fluxes from the powerful and well-studied radio galaxy Cyg A.
151:
152:
153: \subsection {Power of the Cosmic Ray Accelerator}
154:
155: Observations of Cygnus A ($z=0.056$, luminosity distance $\simeq 240 \,\rm Mpc$)
156: by the Chandra X-ray Observatory show that the central
157: $\sim 60~{\rm kpc} \times 120$ kpc size luminous prolate spheroidal region of Cygnus A, called a ``cavity," can be
158: understood as a shock expanding into the accretion cooling-flow gas \citep{wil06}.
159: The kinetic power of expansion of this X-ray cavity deduced by \citet{wil06} is
160: $L_{\rm exp} \approx 1.2\times 10^{46}$ erg s$^{-1}$. This is much larger than the total
161: radio luminosity $L_{\rm lobes}\approx 10^{45}$ erg s$^{-1}$
162: of the two bright radio lobes of Cygnus A \citep{pdc84}.
163: The value of $L_{\rm exp}$ should be considered as a {\it lower}
164: limit to the overall power $L_{CR}$ of cosmic rays
165: injected into the cavity, if the cosmic-ray power is assumed to drive the expansion of the cavity.
166: A few times larger power, $L_{\rm CR}\sim 4\times 10^{46}$ ergs s$^{-1}$
167: therefore seems a reasonable assumption.
168:
169:
170: This value is in a good agreement with the ``neutral beam'' model
171: \citep{ad01,ad03},
172: which explains the collimated relativistic X-ray jets that remain straight up to
173: $\simeq 1$ Mpc distances in sources such as Pictor A \citep{wys01}
174: as the result of energy transport by beams of UHE neutrons and gamma-rays.
175: { These linear X-ray features, surrounded by a broader and less collimated
176: radio structure in Pictor A, and coincident with
177: narrow radio structures exhibiting bends and deflections
178: in Cyg A
179: \citep{car96,sb08},
180: terminate in X-ray hot spots.
181: Because of the large inclination angle, the X-ray jets in Cyg A cannot
182: be detected. However detection of bright X-ray hot spots
183: located at $\simeq 60 \,\rm kpc$ distances on opposite sides
184: of the nucleus \citep{wil06} strongly implies production of
185: collimated X-ray jets also in Cyg A.
186: }
187:
188:
189: Detailed calculations in the framework of this model show that
190: neutral beams of ultrarelativistic neutrons and $\gamma$ rays,
191: produced by the compact relativistic jets in the central sub-parsec scale
192: environment of FRII galaxies, can take a few percent of the total power
193: { of this inner jet. That energy is
194: then deposited, after $\beta$-decay of neutrons
195: $n\rightarrow p+ e +\nu_e$, on distance scales
196: $l_d \simeq E_{20}\,\rm Mpc$, and also via photopair production of UHE
197: gamma-rays through the process $\gamma+\gamma \rightarrow e^- +e^+$.
198: These secondary charged relativistic particles initially form a beam
199: in the same direction as the jet,
200: and can effectively interact with and transfer momentum and energy to the ambient
201: magnetized medium, and thus can be the basic energy
202: reservoir for the jet at large distances from Cyg A.}
203: The total radio luminosity $\simeq 9\times 10^{44}\, \rm erg \, s^{-1}$
204: detected from the lobes of Cyg A \citep{cb96,wil06} imposes the minimum
205: power requirement for the beam. Given the $\sim (2-3)\%$ efficiency
206: of neutral beam production, the acceleration power of the CRs in the
207: relativistic inner jet must be
208: $L_{\rm CR}\gtrsim (3-5)\times 10^{46} \,\rm erg/s$.
209: { This power is then released in CRs when the inner jet
210: decelerates to subrelativistic speeds in the dense medium at kpc
211: scales.}
212: This scenario is in agreement with the assumption that the expansion
213: of the cavity against
214: the cooling flow observed \citep{Smith02}
215: at distances $\lesssim 70$ -- 100 kpc is powered by cosmic-ray pressure.
216:
217:
218: { Protons are accelerated in the inner jet of Cyg A
219: to a maximum energy $E_{max}\approx 10^{20}$ eV, consistent with size scales
220: and magnetic fields inferred from a synchrotron model of blazars corresponding
221: to Cyg A if observed along its jet.
222: This value of $E_{max}$ is in accord with
223: the spatial extent of the X-ray hot spot where the large-scale jet terminates
224: due to the $\beta$-decay of the remaining high-energy neutrons in the beam.
225: At this distance, the injection power from the decaying neutrons is
226: balanced by the ram pressure of the
227: external medium. Thus, for $l_d \sim 100$ kpc-long jets originating from
228: neutral beams,} the inner jets must accelerate protons to
229: $E\gtrsim 10^{19}$eV. In our model, we assume that acceleration of UHECRs
230: to $E\sim 10^{20}$ eV occurs.
231:
232:
233: \subsection {Magnetic Field Strength}
234:
235: { The lower limit to the equipartition magnetic field in the
236: radio lobes of Cyg A inferred from radio observations is $\approx 60\,\rm \mu G$ \citep{car91}.
237: The equipartition magnetic field in the cavity derived from
238: X-ray observations is $B\simeq 120\,\rm \mu G$ \citep{wil06}.}
239: Moreover, the field outside the cavity could also be very high
240: { given the large thermal electron energy density inferred from the
241: X-ray emission at $\lesssim 8^\prime $ or in the
242: $\lesssim 500\,$kpc
243: environment around Cyg A.}
244: The pressure of hot thermal gas decreases from $2.3\times 10^{-10}$ erg cm$^{-3}$ in the regions near
245: the cavity to $\lesssim 10^{-11}$ erg cm$^{-3}$ at distances $\gtrsim 500$
246: kpc from the center of Cyg A \citep{Smith02}.
247: The corresponding equipartition magnetic fields would then vary from
248: $\sim 80\,\rm \mu G$ to $\sim 15\,\rm \mu G$ at the periphery of this region.
249:
250: \subsection{Injection Age}
251:
252: The duration of injection of UHECRs, i.e.,
253: the jet injection age, represents one of the important parameters of the model.
254: \citet{wil06} find that the expansion age of the cavity,
255: determined from the speed of the shock deduced from the analysis of X-ray data,
256: is $t\sim 3\times 10^7\,$yr. The injection age can be larger than this dynamical
257: age because the derived value neglects
258: the magnetic-field pressure of the intracluster medium
259: (ICM) upstream of the shock, as recognized
260: by \citet{wil06}.
261: Also, this age estimate neglects the infall velocity of
262: the cooling flow.
263: For the total mass $M \approx 2\times 10^{13}M_\odot$ enclosed at
264: $r\leq 50\,\rm kpc$ distances \citep{Smith02}, the virial speed
265: $v_{\rm vir} \approx c/\sqrt{r/r_{\rm S}} \cong 2000 \,\rm km/s$,
266: { where $r_{\rm S} = 2GM/c^2 \approx 6\times 10^{18}\,$cm
267: is, formally, the Schwarzschild radius for the total mass $M$.
268: The value of $v_{\rm vir}$ gives a measure of the accretion/cooling flow
269: velocities at radius $r$, and is comparable to the
270: average $\sim 1500 \,\rm km/s$ speed of the shock derived by \citet{wil06}
271: in the rest frame of the fluid upstream of the shock.}
272: %
273: % Although the hot spot advance speed could be faster when Cygnus A was younger, implying
274: % a shorter duration for UHECR injection,
275: These factors can significantly decrease the speed
276: of expansion of the cavity in the stationary frame, so that the real
277: injection age of the cavity can be significantly larger than the age inferred
278: by \citet{wil06}.
279:
280:
281: We now estimate the age of activity of the black-hole jet from energetics
282: arguments. The inferred jet power, $L_{jet} \simeq 4\times 10^{46} \,\rm erg/s$,
283: corresponds to $\simeq 12\%$ of the Eddington luminosity for
284: a black-hole mass $M_{\rm BH} \simeq 2.5 \times 10^{9} M_{\odot}$ in Cygnus A
285: \citep{tud03}.
286: To produce such power, the black hole should accrete mass at the rate
287: $\dot{M}=L_{jet}/\eta c^2= (7/\eta_{-1}) M_\odot \,\rm yr^{-1}$ with an efficiency
288: $\eta = 10^{-1} \eta_{-1}$,
289: { with $\eta_{-1} \approx 1$}.
290: The age of the central black hole {
291: is estimated by its growth time
292: $t\simeq M_{\rm BH}/\dot{M} = 3.6\times 10^{8} \eta_{-1} \,\rm yr$,
293: giving an upper estimate for the jet's age as the jet
294: might be active for only a fraction of the BH growth phase.
295: These estimates are in accord with the characteristic jet age $t\sim 10^8$ yr
296: inferred from a model for ``cocoon''
297: (or cavity) dynamics by \citet{bc89}.}
298:
299: \subsection{Cosmic Ray Diffusive Confinement Time}
300:
301: The maximum confinement timescale $t_{conf}$ of UHECRs in a source of
302: size $r$ is given in the Bohm diffusion approximation by
303: $t_{conf} \cong r^2/2D_{\rm B}$, where the Bohm diffusion
304: coefficient $D_{\rm B} = c r_{\rm L}/3$,
305: { and $r_L \cong E/QB$ is the Larmor radius of a particle with charge $Q$}.
306: From this expression we obtain
307: the UHECR proton confinement time
308: \begin{equation}
309: t_{conf} \cong 0.95\times 10^7 \; {(r/100~{\rm kpc})^2 } E_{20}^{-1}\;(B/20~\mu {\rm G})\; {\rm yr}\;.
310: \label{tesc}
311: \end{equation}
312: For $r\sim 50$ kpc and $B \cong 120~\mu$G, $t_{conf} \cong 15$ Myr for $\approx 10^{20}$ eV
313: protons.
314:
315:
316: \begin{figure}[t]
317: \center
318: \includegraphics[width= 8.0cm]{f1.eps}%12.0cm
319: \caption{Comparison of timescales for energy losses through
320: photomeson (dashed curve)
321: and photopair (dot-dashed curve) production,
322: confinement (solid line) with $B = 20~\mu$G and $r = 500$ kpc,
323: and injection with $t_{age} = 10^8$ yrs
324: (triple-dot-dashed line). The inset shows
325: the fraction $\Lambda(E)$
326: of energy extracted by photohadronic processes from UHECRs with
327: energy $E$.}
328: \vskip0.2in
329: \label{f1}
330: \end{figure}
331:
332:
333:
334: Fig.\ \ref{f1} compares timescales $t_{loss}$ for energy losses due to photomeson and
335: photopair production
336: with values of $t_{conf}$ for characteristic parameters $r = 500$ kpc
337: and $B = 20~\mu$G in the region surrounding Cyg A.
338: %used in Fig.\ \ref{f2} to calculate the $\gamma$-ray fluxes predicted from Cygnus A.
339: Also shown is the value of $t_{age} = 10^8$ yrs for the assumed injection age.
340: The fraction of energy of accelerated protons
341: that could be extracted is given by $\Lambda(E) = \min(t_{conf},t_{loss},t_{age})/t_{loss}$.
342: This figure shows that $\gtrsim 3$\% of the total energy of UHECRs with
343: $E \gtrsim 3\times 10^{18}$ eV
344: can be extracted through photohadronic processes. Because of the increased
345: confinement time and the much
346: lower production threshold for photopair than photomeson processes, the photopair process
347: can make a comparable or dominant contribution to the electromagnetic channel compared to
348: photomeson processes.
349:
350: \section{Results}
351:
352:
353: \begin{figure}[t]
354: \center
355: \includegraphics[width= 8.0cm]{f2.eps}%12.0cm
356: \caption{Fluxes from the electromagnetic cascade initiated
357: within $0.5\,$Mpc distances from the core of Cyg A, assuming
358: an average magnetic field $B = 20~\mu$G in the cavity.
359: The total injection power of UHECR protons
360: is $4\times 10^{46}\, \rm erg/s$. The photon fluxes expected for
361: $t_{age} = 30\,$Myr, 100\,Myr and 300\,Myr are shown by the open dots,
362: full dots and stars, respectively.
363: The solid, dashed, and dot-dashed curves show three generations of
364: synchrotron (heavy curves) and Compton (light curves) cascade radiations
365: in the case of $t=100\,\rm Myr$. Note that the third
366: generation of Compton cascade radiation is too weak to appear on the plot.
367: %Compton fluxes to the total flux for $t_{age} = 100$ Myr.
368: Sensitivities for the FGST and VERITAS are also shown.}
369: \vskip0.2in
370: \label{f2}
371: \end{figure}
372:
373:
374: Fig.\ \ref{f2} shows $\gamma$-ray fluxes calculated for the electromagnetic
375: cascade initiated by the injection of $4\times 10^{46}\,\rm erg$ s$^{-1}$
376: of UHECR protons into the cavity, and further cosmic ray interactions with
377: the EBL in the $r=0.5\,\rm Mpc$ region
378: surrounding Cyg A. Even though the EBL is dominated by the CMBR,
379: interactions with the diffuse infrared/optical radiation are also
380: included in the calculations, using the EBL of \citet{pbs05}.
381: The injection spectrum of UHECR protons is a power-law with an $\alpha
382: = -2.1$ number spectral index with a low-energy cutoff at $E = 1$ TeV and a high-energy
383: exponential cutoff at $E = 3\times 10^{20}$
384: eV. We assume a mean magnetic field $B=20~\mu G$.
385: Escape of particles is given by the Bohm diffusion
386: approximation in a single zone model. The method
387: of calculation follows the approach described by \citet{ad03}.
388:
389: The solid, dashed, and dot-dashed curves in Fig.\ \ref{f2} show the contributions to the
390: fluxes from the photohadronic secondary electrons and the first two generations of
391: cascade electrons for injection ages $t_{age}= 10^8 \,\rm yr$.
392: The secondary electrons also include
393: the electrons from $\pi^0$-decay $\gamma$ rays that convert promptly into
394: electron-positron pairs inside the source. In the strong magnetic field of the
395: Cyg A environment, the Compton flux is dominated by synchrotron radiation
396: from the photohadronic secondaries and the first generation of cascade electrons.
397:
398: The open dots,
399: full dots and stars in Fig.\ \ref{f2} show the received $E \cdot F(E)$ spectral energy fluxes
400: for injection ages
401: $t_{age} = 30\,$Myr, 100\,Myr and 300\,Myr,
402: respectively. The lower and higher energy peaks in the spectral energy distributions
403: primarily result from
404: photopair and photomeson processes, respectively. The sensitivities for a
405: one-year, $5\sigma$ detection of a point source with
406: the FGST in the scanning mode,\footnote{www-glast.slac.stanford.edu/software/IS/glast\_lat\_performance.htm;
407: note that the FGST sensitivity will be somewhat poorer than shown
408: at the $5.76^\circ$ galactic latitude of Cyg A due to Galactic background.} and for a
409: 50 hour,
410: 5$\sigma$ detection with VERITAS are shown. The lower bound of the VERITAS sensitivity applies
411: to a point source, and the upper bound to a source of $25^\prime$ extent.
412:
413: \section{Discussion and Conclusions}
414:
415: If the radio lobes of Cyg A are powered by UHECR production
416: from the inner pc-scale jets, then trapping of these particles in the
417: surrounding strong
418: magnetic-field region leads to the production of secondary $\gamma$ rays that should be
419: significantly detected with the FGST in one year of observation if
420: the jet injection age is $\gtrsim 100$ Myr. If radio galaxies
421: are not the sources of UHECRs, then Cygnus A will not be detected
422: by the FGST.
423: Cyg A could also be detected with VERITAS in a 50 hour pointing, depending
424: on the duration of activity
425: of the central engine and the level of the EBL.
426:
427: Detection of GeV $\gamma$ rays from Cyg A with the FGST might
428: also be expected to arise from other processes.
429: The $\sim 10$ -- 100 GeV radio-emitting electrons
430: from the lobes
431: of radio galaxies will Compton-scatter CMB photons to MeV -- GeV
432: energies \citep[e.g.,][]{che07}. For the strong magnetic field, $\approx 60~\mu$G,
433: in the lobes of Cyg A, however,
434: the ratio of the magnetic-field to CMBR energy densities is
435: $\approx 400$. Thus the total energy flux of
436: Compton-scattered CMBR from Cyg A is $\approx 10^{45}$
437: erg s$^{-1}/[400(4\pi d^2)] \cong 4\times 10^{-13}$ ergs cm$^{-2}$
438: s$^{-1}$, with the $E\cdot F(E)$ flux a factor of $\sim 5$ -- 10 lower.
439: As can be seen from Fig.\ \ref{f2},
440: this process is almost two orders of magnitude below the UHECR-induced synchrotron flux,
441: and well below the FGST sensitivity.
442:
443: {
444: \citet{ias05} considered fluxes expected from the $\lesssim 1 \,\rm Mpc$ halos of clusters of
445: galaxies with weaker magnetic fields, $B\simeq 0.1$-$1\,\mu$ G in a model
446: where acceleration of UHECRs occurs in accretion shocks in the cluster. Because of
447: lower maximum energies of accelerated protons, $E\lesssim 10^{19}\,\rm eV$, and
448: lower magnetic fields, this model predicts hard spectral fluxes of Compton origin
449: peaking at TeV energies}.
450: %
451: \citet{ga05} predicted that synchrotron radiation from
452: $\gtrsim 10^{18}\,\rm eV$ electrons is produced by
453: secondaries of UHECRs that leave the acceleration region
454: and travel nearly rectilinearly through
455: weak intergalactic magnetic fields at
456: the level $B \sim 10^{-7}$ -- $10^{-9}\,$G. These sources would appear
457: as point-like quiescent GeV -- TeV sources with spectra in the
458: GeV domain as hard as $\alpha \cong -1.5$, and very soft,
459: $\alpha \lesssim -3$ spectra in the 100 GeV -- TeV
460: domain.
461:
462: { In contrast to both these models}, we predict
463: soft 0.1 -- 1 GeV spectra with $\alpha \cong -2.5$ and hard, $\alpha \cong -2$
464: spectra at TeV energies due to the much
465: higher magnetic field in the confinement region. These models can
466: be clearly distinguished if Cygnus A is resolved by the Fermi Gamma ray Space
467: Telescope or the ground-based
468: $\gamma$-ray telescopes, as the emission region in our model subtends an angle
469: $\approx 10^\prime$.
470:
471: Because Cygnus A lies outside the GZK horizon, only UHECRs with energy
472: below the GZK energy could be correlated with this source. Other closer
473: FRII radio galaxies that are correlated with the arrival directions of
474: UHECRs are, however, candidate sources of $\gamma$ rays made through
475: the mechanism proposed here. IGR J21247+5058 at $z = 0.02$ or $d\approx 80$
476: Mpc, recently discovered with INTEGRAL \citep{mol07}, is 2.1 degrees
477: away from a HiRes Stereo event with $E > 56$ EeV (C.\ C.\ Cheung, private
478: communication, 2008).\footnote{This radio galaxy was identified as such only
479: recently and would not have appeared in the list of AGN used by the HiRes
480: collaboration in their correlation study \citep{abb08}.} PKS 2158-380 at
481: $\approx 140$ Mpc is also within 3.2$^\circ$ degrees of an Auger UHECR with
482: $E> 57$ EeV \citep{mos08}. By comparison with Cyg A, these are low luminosity
483: FRIIs, and their predicted flux level will require detailed modeling for each
484: source, as done here for Cyg A.
485: Variability of $\gamma$-ray emission would rule out our model.
486:
487: % and imply
488: %instead a model of misaligned jet emission. This could be the case, for example,
489: %if the $\gamma$-ray emission detected from Centaurus A with EGRET \citep{sre99} is variable.
490: %If, instead, $\gamma$-ray emission from Cen A is steady, then
491: % a confinement scenario as studied here could
492: %also apply to this source. A more detailed study of this possibility
493: % is considered in separate work.
494:
495:
496: \acknowledgments
497:
498: We thank Vladimir Vassiliev for discussion and providing the VERITAS sensitivities,
499: Teddy Cheung and Felix Aharonian for important comments, and the referee for a detailed
500: report.
501: The work of AA and visits to NRL were supported by the
502: GLAST Interdisciplinary Science Investigation Grant DPR-S-1563-Y.
503: The work of CDD is supported by the Office of Naval Research.
504:
505: \begin{thebibliography}{}
506:
507: \bibitem[Abbasi et al.(2008)]{abb08} Abbasi, R.~U., et al.\
508: 2008, ArXiv e-prints, 804, arXiv:0804.0382
509:
510: \bibitem[Aharonian et al.(1994)]{acv94}
511: Aharonian, F.~A., Coppi, P.~S., V\"olk, H.~J.\ 1994, ApJ, 423, L5
512:
513: \bibitem[Aharonian(2002)]{aha02}
514: Aharonian, F.~A.\ 2002, MNRAS, 332, 202
515:
516: \bibitem[Atoyan \& Aharonian(1999)]{aa99} Atoyan, A.~M., \&
517: Aharonian, F.~A.\ 1999, MNRAS, 302, 253
518:
519: \bibitem[Atoyan \& Dermer(2001)]{ad01}
520: Atoyan, A., and Dermer, C.\ 2001, PRL 87, 221102
521:
522: \bibitem[Atoyan \& Dermer(2003)]{ad03}
523: Atoyan, A., and Dermer, C.\ 2003, ApJ, 586, 79
524:
525: %\bibitem[Armengaud et al.(2006)]{asm06} Armengaud, E., Sigl,
526: %G., \& Miniati, F.\ 2006, \prd, 73, 083008
527:
528:
529: \bibitem[Begelman
530: \& Cioffi(1989)]{bc89} Begelman, M.~C., \& Cioffi, D.~F.\ 1989, \apjl, 345, L21
531:
532: \bibitem[Carilli et al.(1996)]{car96} Carilli, C., Perley,
533: R., Bartel, N., \& Dreher, J.\ 1996, in Cygnus A -- Study of a Radio Galaxy,
534: eds.\ by C.L. Carilli and D.E. Harris (Cambridge: Cambridge University Press), 76
535:
536: \bibitem[Carilli et al.(1991)]{car91} Carilli, C.~L., Perley,
537: R.~A., Dreher, J.~W., \& Leahy, J.~P.\ 1991, \apj, 383, 554
538:
539: \bibitem[Carilli \& Barthel(1996)]{cb96} Carilli, C.~L., \& Barthel, P.~D.\ 1996, \aapr, 7, 1
540:
541: \bibitem[Cheung(2007)]{che07} Cheung, C.~C.\ 2007, The First
542: GLAST Symposium, S.\ Ritz, P.\ Michelson, \& C.\ Meegan, eds.\ (AIP: New York) 921, 325
543:
544: % \bibitem[Coppi \& Aharonian(1997)]{ca97} Coppi, P.~S., \& Aharonian, F.~A.\ 1997, \apjl, 487, L9
545:
546:
547: \bibitem[ Dermer \& Atoyan(2004)]{da04}
548: Dermer, C.~D., \& Atoyan, A.\ 2004, ApJ, 611, L9
549:
550:
551: \bibitem[Feretti et al.(1995)]{fer95}
552: Feretti, L., Dallacasa, D., Giovannini, G., \& Tagliani, A.\ 1995, A\&A,
553: 302, 680
554:
555: \bibitem[Ferrari et al.(2008)]{fer08} Ferrari, C., Govoni,
556: F., Schindler, S., Bykov, A.~M.,
557: \& Rephaeli, Y.\ 2008, Space Science Reviews, 134, 93
558:
559: \bibitem[Gabici \& Aharonian(2005)]{ga05}
560: Gabici, S., \& Aharonian, F. A., 2005, Phys. Rev. Lett. 95, 251102
561:
562: \bibitem[Halzen \& Hooper(2002)]{hh02}
563: Halzen, F., \& Hooper, D.\ 2002, Reports of Progress in Physics, 65, 1025
564:
565: \bibitem[Harari et al.(2006)]{hmr06} Harari, D., Mollerach,
566: S., \& Roulet, E.\ 2006, J.\ Cosmology and Astroparticle Physics, 11, 12
567:
568: \bibitem[Inoue et al.(2005)]{ias05}
569: Inoue, S., Aharonian, F.~A., and Sugiyama, N. 2005, \apjl, 628, L9
570:
571: \bibitem[Kim et al.(1990)]{kim90}
572: Kim, K.-T., Kronberg, P. P., Dewdney, P. E., \& Landecker, T. L.\ 1990,
573: ApJ, 355, 29
574:
575: \bibitem[Molina et al.(2007)]{mol07} Molina, M., et al.\
576: 2007, \mnras, 382, 937
577:
578: \bibitem[Moskalenko et al.(2008)]{mos08} Moskalenko, I.~V.,
579: Stawarz, L., Porter, T.~A.,
580: \& Cheung, C.~C.\ 2008, ArXiv e-prints, 805, arXiv:0805.1260
581:
582: \bibitem[M\"ucke et al.(1999)]{muc99}
583: M\"ucke, Rachen, J.P., A., Engel, R., Protheroe, R.J., and Stanev, T.
584: 1999, Pub.\ Astron.\ Soc.\ Australia, 16, 160
585:
586: \bibitem[Perley et al.(1984)]{pdc84} Perley, R.~A., Dreher,
587: J.~W., \& Cowan, J.~J.\ 1984, \apjl, 285, L35
588:
589:
590: \bibitem[Pierre Auger Collaboration(2007)]{Auger07}
591: The Pierre Auger Collaboration, et al., 2007, Science, 318, 938
592:
593: \bibitem[Primack et al.(2005)]{pbs05} Primack, J.~R.,
594: Bullock, J.~S., \& Somerville, R.~S.\ 2005, High Energy Gamma-Ray Astronomy, 745, 23
595:
596: \bibitem[Scheuer(1974)]{sch74} Scheuer, P.~A.~G.\ 1974,
597: \mnras, 166, 513
598:
599: \bibitem[Smith et al.(2002)]{Smith02} Smith, D.~A., Wilson,
600: A.~S., Arnaud, K.~A., Terashima, Y., \& Young, A.~J.\ 2002, \apj, 565, 195
601:
602:
603: %\bibitem[Sreekumar et al.(1999)]{sre99} Sreekumar, P.,
604: %Bertsch, D.~L., Hartman, R.~C., Nolan, P.~L.,
605: %\& Thompson, D.~J.\ 1999, Astroparticle Physics, 11, 221
606:
607: \bibitem[Steenbrugge \& Blundell(2008)]{sb08} Steenbrugge, K.~C., \& Blundell, K.~M.\ 2008, \mnras, 388, 1457
608:
609:
610: \bibitem[Tadhunter et al.(2003)]{tud03} Tadhunter, C.,
611: Marconi, A., Axon, D., Wills, K., Robinson, T.~G.,
612: \& Jackson, N.\ 2003, \mnras, 342, 861
613:
614: \bibitem[Wehrle et al.(1998)]{weh98}
615: Wehrle, A.\ E.\ et al.\ 1998, ApJ, 497, 178
616: %\bibitem[Wick et al.(2004)]{wda04} Wick, S.~D., Dermer,
617: %C.~D., \& Atoyan, A.\ 2004, Astroparticle Physics, 21, 125
618:
619:
620: \bibitem[Wilson, Young, and Shopbell(2001)]{wys01}
621: Wilson, A. S., Young, A. J., and Shopbell, P. L. 2001, ApJ, 547, 740
622:
623: \bibitem[Wilson et al.(2006)]{wil06}
624: Wilson, A., et al. 2006, ApJ, 644, L9
625: \end{thebibliography}
626:
627: \end{document}