1: \documentclass[12pt,preprint]{aastex}
2: % \documentclass[manuscript]{aastex}
3: % \documentclass[preprint2]{aastex}
4: % \documentclass{emulateapj}
5: % \documentclass[onecolumn]{emulateapj}
6: % \usepackage{epsf}
7:
8: \slugcomment{draft of \today}
9: \shorttitle{Propagation of UHE protons}
10: \shortauthors{Das {\it et al.~}}
11:
12: \def\etal{{\it et al.}}
13: \def\eg{{\it e.g.,}}
14: \def\ie{{\it i.e.,~}}
15:
16: \def\kms{~{\rm km~s^{-1}}}
17: \def\cm3{~{\rm cm^{-3}}}
18: \def\yrs{~{\rm yrs}}
19: \def\Mpc{~h^{-1}{\rm Mpc}}
20: \def\kpc{~h^{-1}{\rm kpc}}
21:
22: \begin{document}
23: \title{Propagation of Ultra-High-Energy Protons through the Magnetized Cosmic Web}
24:
25: \author{Santabrata Das\altaffilmark{1},
26: Hyesung Kang\altaffilmark{2,3},
27: Dongsu Ryu\altaffilmark{4},
28: and Jungyeon Cho\altaffilmark{4}}
29:
30: \altaffiltext{1}
31: {Astrophysical
32: Research Center for the Structure and Evolution of the Cosmos,
33: Sejong University, Seoul 143-747, South Korea: sbdas@canopus.cnu.ac.kr}
34: \altaffiltext{2}
35: {Department of Earth Sciences, Pusan National University, Pusan 609-735,
36: South Korea:\\ kang@uju.es.pusan.ac.kr}
37: \altaffiltext{3}
38: {Author to whom correspondence should be addressed}
39: \altaffiltext{4}
40: {Department of Astronomy and Space Science, Chungnam National University,
41: Daejeon 305-764, South Korea: ryu@canopus.cnu.ac.kr, cho@canopus.cnu.ac.kr}
42:
43: \begin{abstract}
44: If ultra-high-energy cosmic rays (UHECRs) originate from extragalactic
45: sources, understanding the propagation of charged particles through
46: the magnetized large scale structure (LSS) of the universe is crucial
47: in the search for the astrophysical accelerators.
48: Based on a novel model of the turbulence dynamo, we estimate the
49: intergalactic magnetic fields (IGMFs) in cosmological simulations
50: of the formation of the LSS.
51: Under the premise that the sources of UHECRs are strongly associated
52: with the LSS, we consider a model in which protons with $E \geq 10^{19}$
53: eV are injected by sources that represent active galactic nuclei
54: located inside clusters of galaxies.
55: With the model IGMFs, we then follow the trajectories of the protons,
56: while taking into account the energy losses due to interactions with
57: the cosmic background radiation.
58: For observers located inside groups of galaxies like ours,
59: about 70\% and 35\% of UHECR events above 60 EeV arrive within $\sim 15^{\circ}$
60: and $\sim 5^{\circ}$, respectively, of the source position
61: with time delays of less than $\sim 10^7$ yr.
62: This implies that the arrival direction of super-GZK protons might
63: exhibit a correlation with the distribution of cosmological sources
64: on the sky.
65: In this model, nearby sources (within $10 - 20$ Mpc) should contribute
66: significantly to the particle flux above $\sim10^{20}$ eV.
67: \end{abstract}
68:
69: \keywords{cosmic rays -- large scale structure of the universe --
70: magnetic fields -- methods: numerical}
71: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
72: \section{Introduction}
73:
74: Over the past several decades,
75: significant progress has been made on both theoretical
76: and observational fronts in understanding the nature
77: and origin of ultrahigh energy cosmic rays (UHECRs),
78: those with $E \ga 1$ EeV ($ = 10^{18}$ eV)
79: \citep[for recent reviews, see][]{nagano00,bgg06}.
80: Yet, the acceleration mechanism and the corresponding astrophysical
81: ``accelerators" of these energetic particles are still largely unknown.
82: Observational data from several experiments such as the High Resolution
83: Fly's Eye (HiRes) indicate that the mass
84: composition of UHECRs becomes lighter at higher energies \citep{abbasi05}.
85: However, composition analyses that include high energy interactions are
86: often model-dependent and inconclusive \citep{watson06}.
87: According to a recent report from the Pierre Auger Observatory,
88: the mass composition is likely mixed,
89: possibly becoming heavier above 30 EeV \citep{unger07}.
90: The overall distribution of UHECR arrival directions is
91: considered to be consistent with isotropy \citep{bo03}.
92: Exceptions to this general isotropy include the small-scale clusterings
93: of doublets and triplets found in data collected by the Akeno Giant
94: Air Shower Array (AGASA) experiment \citep{takeda99,uchihori00}.
95: In addition, a possible correlation of AGASA events with BL Lacertae objects
96: has been suggested \citep{tinyakov01}.
97: But these claims have been confirmed neither by HiRes \citep{abbasi06}
98: nor by Auger \citep{harari07,armen07}.
99: However, the Auger Collaboration recently reported that
100: the arrival directions of UHECRs above $60$EeV in their data show
101: a correlation with the position of active galactic nuclei (AGNs)
102: lying within 75 Mpc \citep{augerScience}.
103:
104: Since protons with $E \ga 1$ EeV cannot be confined within the
105: Galactic plane, UHECRs likely originate from extragalactic sources.
106: In particular, the overall isotropy of arrival directions
107: suggests that there may be a large number of sources distributed
108: over cosmological distances \citep{nagano00,bo03}.
109: During their propagation through intergalactic space, such protons will lose
110: energy by means of pion and pair production processes while
111: interacting with the cosmic background radiation \citep{g66,zk66}.
112: The flux of ultra-high-energy (UHE) protons from cosmological sources is thus
113: expected to be strongly attenuated,
114: resulting in a significant suppression in the observed spectrum above
115: the GZK threshold energy, $E_{\rm GZK}\approx 40$ EeV \citep{bgg06}.
116: Although the AGASA data show no indication of the GZK suppression
117: \citep{nagano00},
118: both the Yakutsk Extensive Air Shower array and HiRes have reported a suppression of flux above
119: $E_{\rm GZK}$, contradicting the AGASA finding \citep{pravdin99,zech04}.
120: Indeed the same suppression was seen in a recent Auger measurement
121: \citep{facal07}, which seems to have ended the
122: controversy over the presence of the GZK cutoff.
123: The so-called dip-calibrated UHECR spectra from different
124: experiments compiled by \citet{bgg06},
125: appear to be in good agreement with each other
126: and to be consistent with GZK suppression.
127: However, it has yet to be understood whether this suppression is actually due to the
128: GZK cutoff or due to the maximum acceleration energy,
129: $E_{\rm max}$, of astrophysical accelerators.
130: If UHECRs are protons,
131: the GZK energy loss should operate at acceleration sites as well,
132: leading to an $E_{\rm max}$ close to $E_{\rm GZK}$ \citep[see, \eg][]{krb97}.
133:
134: Most UHE protons observed above the GZK energy must come from
135: within the so-called GZK sphere of radius $R_{\rm GZK} \sim 100 $ Mpc, although
136: the proton interaction length at $E_{\rm GZK}$ is $l_{\rm 40EeV}
137: \approx 1$ Gpc, corresponding to $z\sim 0.2$ \citep{berezinsky88}.
138: However, finding cosmological sources inside the GZK sphere from the
139: arrival directions of UHECRs is not straightforward,
140: since their paths are deflected by intergalactic magnetic fields
141: (IGMFs) \citep[e.g.][]{sigletal03,dolagetal04}.
142: Sigl and collaborators have extensively studied the propagation of UHECRs
143: in a structured and magnetized universe, adopting a numerical model for
144: the IGMFs \citep{sigletal03,sigletal04,armen05}.
145: In this model, the IGMFs are generated by means of the Biermann ``battery''
146: mechanism at shocks and then evolved passively in a cosmological hydrodynamic
147: simulation \citep{kulsrude97,rkb98}.
148: The strength of the resulting fields is rescaled to match the
149: simulated field strength in Coma-like clusters to the observed
150: strength, which is on the order of microgauss.
151: UHECRs with $E\ge10$EeV are then injected at cosmological sources.
152: These particles propagate through the magnetized large scale structure
153: (LSS) of the universe, and arrive at a mock observer with deflection
154: angle $\theta$, the angle between the arrival direction and the
155: source's location on the sky.
156: Sigl \etal\ found that the deflection due to IGMFs is significant,
157: with $\theta \ga 20^{\circ}$ above 100 EeV.
158: On the other hand, Dolag \etal\ adopted an IGMF model from a
159: ``constrained'' cosmological simulation employing a magnetohydrodynamic (MHD)
160: version of a smoothed particle hydrodynamics (SPH) code.
161: They found the deflection angle of protons with 100 EeV
162: to be less than $1^{\circ}$, contradicting the estimate by
163: Sigl \etal\ \citep{dolagetal04,dolagetal05}.
164:
165: This controversy over the predicted deflection angles demonstrates
166: the importance of modeling the IGMFs in identifying
167: the astrophysical sources and studying the origin of UHECRs.
168: In order to reexamine this issue, here we adopt IGMFs,
169: based on a novel models of the turbulence dynamo \citet{ryuetal08}
170: In this model, the strength of the magnetic fields is
171: estimated from the local vorticity and turbulent kinetic energy
172: in cosmological structure formation simulations.
173: For the field direction, the passive fields from these simulations are used.
174:
175: The maximum energy of nuclei of charge $Z$ that can be confined and
176: accelerated by astrophysical sources is given by
177: \begin{equation}
178: E_{\rm max} \approx Z \left({V \over c}\right)
179: \left({B \over \mu{\rm G}}\right)
180: \left({L \over {\rm kpc}}\right) 10^3 {\rm EeV},
181: \end{equation}
182: where $V$, $B$, and $L$ are the characteristic flow speed, magnetic
183: field strength, and linear size of the accelerator, respectively \citep{hillas84}.
184: There are a few viable candidates that can produce the required
185: $E_{\rm max} \sim 100$ EeV: jets from AGNs \citep[\eg][]{biermann87},
186: gamma-ray bursts \citep[\eg][]{waxman95}, and cosmological shocks
187: \citep{krj96,krb97}.
188: In this study, we consider AGNs inside galaxy clusters as the sources
189: of UHECRs.
190: Thus, the source position is in effect correlated with the LSS of
191: the universe.
192: Protons with $E\ge 10$EeV are injected at the sources and travel
193: through the simulated magnetized space until they lose energy down to
194: 10 EeV, meanwhile visiting mock observers placed inside groups of galaxies.
195: In this study we focus mainly on the deflection angle, the time delay
196: relative to rectilinear flight, and the energy spectrum of the UHE protons.
197:
198: In the next section, we describe our model IGMFs, cosmic-ray sources, and
199: observers, and the simulations of the propagation of UHE protons
200: in intergalactic space.
201: The results are presented in \S 3.
202: Finally, we conclude in \S 4.
203:
204: \section{Models and Methods}
205:
206: \subsection{Intergalactic Magnetic Fields}
207:
208: Our cosmological hydrodynamic simulations of a concordance $\Lambda$CDM
209: universe are carried out with the following parameters:
210: $\Omega_{\rm BM}=0.043$, $\Omega_{\rm DM}=0.227$, and
211: $\Omega_{\Lambda}=0.73$, $h \equiv H_0$/(100 km/s/Mpc) = 0.7, and
212: $\sigma_8 = 0.8$.
213: A cubic box of comoving size $100 \Mpc$ is simulated using $512^3$
214: grid zones for gas and gravity and $256^3$ particles for dark matter.
215: In the simulation, magnetic fields are generated through the Biermann battery
216: mechanism at structure formation shocks and then evolved passively with the
217: flow motions \citep{kulsrude97,rkb98}.
218: The simulation was repeated for six different realizations of the initial
219: conditions to examine the effects of cosmic variance.
220: Non gravitational effects, including radiative cooling,
221: photo ionization and heating, and feedback from star formation are ignored.
222: Those processes affect the generation and evolution of magnetic fields
223: mainly on small scales \citep{krco07}, which should not alter the
224: large-scale fields primarily responsible for the deflection of UHECRs.
225:
226: In principle, if we were to perform full MHD simulations, we could follow
227: the growth of the IGMFs through stretching, twisting, and folding of
228: field lines, the process known as the turbulence dynamo.
229: In practice, however, the computational resources currently available
230: do not allow high enough numerical resolution to reproduce the full
231: development of MHD turbulence:
232: since the numerical resistivity is larger than the physical resistivity
233: by many orders of magnitude, the growth of the magnetic fields saturates
234: before the dynamo action becomes fully operative \citep{kulsrude97}.
235: So, \citet{sigletal03,sigletal04} rescaled
236: the strength of their passively evolved fields
237: in the postprocessing analysis
238: to match the observed field strength in clusters of galaxies.
239: This rescaling, which hinges on the observed field strength in
240: the intracluster medium, does not necessarily result in correct
241: field strengths for filaments, sheets, and voids.
242: On the other hand, in the MHD SPH simulations of \citet{dolagetal04,
243: dolagetal05}, the initial field strength was adjusted to obtain
244: a microgauss level in clusters of galaxies at the present epoch.
245: They demonstrated that their simulated cluster fields are consistent
246: with various observations, such as rotation measure profiles and
247: the total radio powers of cluster halos.
248: A lack of observations, however, prevents their simulated fields in
249: low-density regions from being tested against the real magnetic fields
250: in filaments and sheets.
251: In the MHD SPH simulations, the flow motions can be resolved
252: reasonably well in high-density regions, where the smoothing length is
253: sufficiently small, ensuring adequate growth of magnetic fields.
254: However, turbulence may not be fully realized in the low-density
255: regions, which have large smoothing lengths, so the field strength in
256: filaments and sheets may be underestimated in such simulations.
257:
258: In this study, we take a new approach detailed in \citet{ryuetal08}.
259: If one assumes that magnetic fields grow as a result of the turbulence dynamo,
260: their energy density can be estimated from the eddy turnover number and
261: turbulent energy density as follows :
262: \begin{equation}
263: \varepsilon_B = \phi \left({t \over t_{\rm eddy}}\right)
264: \varepsilon_{\rm turb}.
265: \end{equation}
266: Here the eddy turnover time is defined as the reciprocal of the vorticity
267: at driving scales, $t_{\rm eddy} \equiv 1/\omega_{\rm driv}$
268: (${\vec \omega} \equiv {\vec\nabla}\times{\vec v}$), and $\phi$ is the
269: conversion factor from turbulent to magnetic energy, which
270: is determined from high-resolution MHD simulations of turbulence.
271: For our model IGMFs, the number of eddy turnovers is estimated as
272: the age of the universe multiplied by the magnitude of the local vorticity,
273: that is, $t_{\rm age}\omega$.
274: The local vorticity and turbulent energy density are calculated
275: from the cosmological structure formation simulations.
276: The energy density given by equation (2) fixes the strength of the IGMFs,
277: so our model requires neither rescaling of the field strength
278: nor adjustment of the initial fields.
279: As in the work of Sigl \etal, we assume that the topology of the IGMFs in the LSS
280: can be represented statistically by the topology of the passive
281: magnetic fields in the cosmological simulations.
282:
283: Figure 1 shows a two-dimensional slice of the magnetic field strength
284: in our model at the present epoch.
285: The IGMFs are structured and well correlated with the weblike cosmic
286: distribution of matter.
287: The strongest magnetic fields, with $B \ga 0.1\mu$G, are found inside
288: and around clusters, while the fields are weaker in filaments,
289: sheets, and voids.
290: Overall, there is a correlation between field strength and gas
291: density, as can be seen in Figure 2({\it left}).
292: In the regions of galaxy clusters, with
293: $\rho_{\rm gas}/\langle \rho_{\rm gas} \rangle \ga 10^3$,
294: we find $\langle B \rangle \sim 1~\mu$G.
295: For typical filamentary regions, with
296: $\rho_{\rm gas}/\langle \rho_{\rm gas} \rangle \sim 10$,
297: the field has $\langle B \rangle \sim 10^{-8}$ G.
298: By comparison, the average field strength in filaments is found to be
299: $\langle B \rangle \sim 10^{-7}$ G by \citet{sigletal04} and
300: $\langle B \rangle \sim 10^{-10}$ G by \citet{dolagetal05}.
301: The right panel of Figure 2 shows the volume fraction, $df/d \log B$
302: ({\it solid line}), and its cumulative distribution, $f(>B)$ ({\it dotted line}) and $f(<B)$
303: ({\it dot-dashed line}).
304: In our model, the volume filling factor for $B >10^{-8}~{\rm G}$
305: is $f(>10^{-8}{\rm G}) \approx 0.01$.
306: By comparison, $f(>10^{-8}~{\rm G}) \approx 0.1$ in \citet{sigletal04}
307: and $f(>10^{-8}~{\rm G}) \approx 10^{-4}$ in \citet{dolagetal05}.
308: Hence, the field strength in filaments in our model lies
309: in the middle of the values from the models used by these two groups,
310: that is, lower than Sigl \etal's but higher than that of Dolag \etal's.
311:
312: \subsection{AGNs as UHECR Sources}
313:
314: In the framework of bottom-up acceleration models, AGNs are the most
315: studied candidate astrophysical accelerators to produce
316: cosmic-ray nuclei beyond the GZK energy \citep[for a review, see ][]{bgg06}.
317: As noted in \S I, it has also been suggested that cosmological shocks
318: may accelerate nuclei of charge $Z$ up to an $E_{\rm max}$ of
319: a few times $10^{19}\ Z$ eV, but it is unlikely that protons would be
320: accelerated beyond the GZK energy by such shocks \citep{krb97, inoue07}.
321: However, if some UHECRs are iron nuclei, cosmological shocks can
322: provide the acceleration sites for super-GZK cosmic rays \citep{inoue07}.
323: In this paper, we consider only AGN-like objects as sources
324: of UHECRs, since we focus on the propagation of UHE protons.
325:
326: We identify X-ray clusters characterized by the X-ray emission weighted gas temperature
327: $kT>1$ keV in the cosmological simulations as source locations.
328: One AGN is placed at the center of each ``host'' cluster.
329: With this selection criterion, the source locations are chosen, in effect,
330: at high-density regions with the strongest field strength.
331: Table 1 shows the number of such clusters found in the six cosmological
332: simulations ($\Lambda$CDM1-$\Lambda$CDM6) with different initial
333: conditions.
334: The three-dimensional distribution of the 18 sources in $\Lambda$CDM1
335: is shown in Figure 3.
336: Given the simulation volume of $(100 \Mpc)^3$, the mean separation of
337: sources is $l_s \approx 40 \Mpc$ and the source number density is
338: $n_s = 2-3 \times10^{-5} h^{3}{\rm Mpc}^{-3}$,
339: which is consistent with the required UHECR source density
340: inferred from the small-scale clustering found in the AGASA data
341: \citep{Yoshiguchietal03, sigletal03, bm04}.
342: The field strength at the source locations mostly lies in the range
343: $ 0.1 \mu {\rm G} \la B_s \la 2 \mu {\rm G}$, with a peak at $\sim 1 \mu$G;
344: its distribution is shown in Figure 4 ({\it left}).
345:
346: \subsection{Groups of Galaxies as Mock Observers}
347:
348: The key physical condition for an ``observer'' that is most relevant to
349: this study is the strength and direction of the magnetic fields, since we
350: are interested primarily in the deflection angles and time delays of
351: UHECRs.
352: Little is known about the magnetic fields in the intergalactic space
353: within the Local Group.
354: So, we select groups of galaxies identified in the simulation data that
355: have similar halo gas temperatures to the Local Group,
356: that is, $ 0.05 {\rm keV} < kT < 0.5 {\rm keV}$ \citep{rasmus01},
357: assuming that these groups are located in magnetic environment
358: similar to that of the Local Group.
359: There are about $1000-1400$ identified groups with gas temperatures
360: in this range inside the simulation volume (see Table 1).
361: As can be seen in Figure 3, these groups are not distributed uniformly
362: but are located mostly along filaments, following the matter distribution
363: of the LSS.
364: A mock observer, modeled as a sphere of radius $R_{\rm obs}=0.5 \Mpc$,
365: is placed at each group.
366: The value of $R_{\rm obs}$ is chosen so that the observer's sphere
367: is well contained within the associated filament, since the typical
368: thickness of the magnetized region around a filament is about
369: $2-3 \Mpc$.
370: If we were to use an $R_{\rm obs}$ smaller than the value adopted,
371: smaller cross sections would lead to smaller detection rates of cosmic rays
372: in our numerical experiment described below.
373: The right panel of Figure 4 shows the field strength within the
374: observer spheres for the six cosmological simulations:
375: $ 10^{-4} {\mu \rm G} \la B_{\rm obs} \la 0.1 {\mu \rm G}$ with a peak at
376: $\sim 10^{-9}$ G.
377: This illustrates the distribution of magnetic field strength in filaments.
378:
379: \subsection{Propagation of UHE protons in the Intergalactic Space}
380:
381: The characteristics of the particle spectrum accelerated at UHECR
382: sources are largely unknown.
383: We thus make the simple assumption that the accelerated protons have
384: a power-law energy spectrum at their source: $N_{\rm inj}(E_{\rm inj})
385: \propto E_{\rm inj}^{-\gamma}$ for 10 EeV $\le E_{\rm inj} \le 10^3$ EeV.
386: In practice, $\gamma=0$ is used to generate a flat injection spectrum
387: at the sources, and later a weighting factor proportional to
388: $E_{\rm inj}^{-\gamma}$ is applied to the statistics (except for the distribution
389: in Fig. 5; see \S3.1for details).
390: For each $\Lambda$CDM simulation,
391: a total of $3\times10^4$ particles are randomly distributed over the sources
392: and then launched in random directions
393: from random positions inside a sphere of radius $0.5\Mpc$.
394:
395: We follow the trajectories of the UHE protons by numerically integrating
396: the equations of motion in our model IGMFs,
397: \begin{equation}
398: \frac{d\vec{r}}{dt}=\vec{v} ~ ; ~~~~ \frac{d\vec{v}}{dt}=\frac{Ze}{mc}
399: \left(\vec{v}\times \vec{B} \right).
400: \end{equation}
401: The energy losses due to photo pion production and pair production
402: are treated with the continuous loss approximation \citep{bgg06},
403: but the adiabatic losses due to cosmic expansion are ignored,
404: because the largest source-to-observer distance is $D_{\rm max} \sim 1$
405: Gpc in our experiment, corresponding to only $z_{\rm max} \approx 0.2$.
406: In practice, the distances are not known in advance, since we are integrating
407: the trajectory from sources to observers.
408:
409: The simulation box of $(100 \Mpc)^3$ at $z=0$ is used to
410: define sources (host clusters), mock observers (groups), and the
411: IGMF data.
412: Additional virtual boxes with the same distribution of mock
413: observers and IGMF data are periodically stacked indefinitely.
414: Particles are injected from the sources only in the original box.
415: Then they travel through the magnetized space consisting
416: of the original box and the replicated periodic boxes.
417: Once a particle visits an observer sphere,
418: the arrival direction, time delay and energy of the particle are
419: registered as a ``recorded event.''
420: We let the particle continue its journey, visiting several observers
421: during its full flight, until its energy falls to 10 EeV.
422: With $3\times 10^4$ protons injected for each simulation box, about
423: $2.6 \times 10^5$ events are recorded in total for all six
424: $\Lambda$CDM simulations.
425:
426: In our propagation experiment, the source-to-observer distance $D$
427: can be arbitrarily small because the specific way that we set up source
428: and observer locations, and the statistics of recorded events depend
429: on the minimum value of $D$.
430: In reality, the closest AGNs to us are Centaurus A, at 3.42 Mpc
431: \citep{ferrarese07}, in the southern hemisphere and M87, at 16.7 Mpc
432: \citep{mei07}, in the northern hemisphere.
433: So we mostly present the results for $D_{\rm min}=3$ Mpc or $D_{\rm min}=10$ Mpc.
434: The recorded events with $D < D_{\rm min}$ are excluded from the analysis.
435: For $D_{\rm min} \ge 10$ Mpc, however, the results become less sensitive
436: to the value of $D_{\rm min}$.
437:
438: \section{Results}
439:
440: \subsection{Deflection Angle and Time Delay of UHECRs}
441:
442: With a gyroradius
443: \begin{equation}
444: r_g = 10\ {\rm kpc}\left({E \over 10^{19}{\rm eV}}\right)
445: \left({B \over \mu{\rm G}}\right)^{-1},
446: \end{equation}
447: UHE protons will suffer significant deflection during their propagation
448: when they pass regions with $B \ga 10^{-8}$ G, that is, clusters and filaments.
449: Here the filamentary regions are more significant players than clusters, since
450: the volume filling factor of filaments is much larger than that of
451: clusters (see Fig. 2).
452: As a result of this deflection, the actual path traveled by the UHECRs in the
453: presence of the IGMFs can be much longer than a rectilinear path,
454: causing a significant time delay.
455: We therefore measure the deflection angle as the angle between the arrival
456: direction of the cosmic rays and the source position on the sky,
457: and the time delay as the difference between the arrival time and
458: the rectilinear travel time.
459:
460: UHE protons can be also deflected inside the host clusters of sources,
461: before they escape to intergalactic space.
462: Typical clusters have a magnetized core envelope structure
463: with $B_{\rm core} \sim 1\mu$G, $R_{\rm core} \la 0.5 \Mpc$ and
464: $B_{\rm env} \sim 0.01-0.1 \mu$G, extending out to $R_{\rm env}\sim
465: 3 \Mpc$ (see Fig. 1).
466: So, the protons with $E \la E_{\rm GZK}$ injected by AGNs are
467: scattered by turbulent magnetic fields inside the host clusters
468: and confined within the magnetized structure for a while.
469: The scatterings by the turbulent fields local to the sources alone
470: can cause a deflection angle
471: \begin{equation}
472: \theta_{\rm source} \sim \tan^{-1}\left({{\rm a~few~Mpc} \over D}\right),
473: \end{equation}
474: for protons with $E \la E_{\rm GZK}$.
475:
476: Figure 5 shows the distributions of the deflection angle $\theta$
477: and the time delay $t_d$ as functions of the distance $D$ for the
478: $\Lambda$CDM1 simulation.
479: The events are divided into three channels in observed energy,
480: as follows: $10<E_{\rm obs}<30$ ({\it top}), $30<E_{\rm obs}<60$
481: ({\it middle}), and $E_{\rm obs}>60$ ({\it bottom}), where the
482: particle energy is given in units of EeV.
483: The data points are color-coded by the injection energy in the same
484: three channels, that is, red for $10<E_{\rm inj}<30$, blue
485: for $30<E_{\rm inj}<60$, and green for $E_{\rm inj}>60$.
486: For these plots, the calculation was performed with an injection spectrum
487: with $\gamma=2.7$ instead of the flat spectrum, since the recorded
488: data points in this type of representation cannot be weighted with
489: a factor proportional to $E_{\rm inj}^{-\gamma}$.
490:
491: Sub-GZK protons, with $E_{\rm obs}< 60$ EeV, come from sources
492: as distant as $D\sim 1$ Gpc.
493: For these particles, the distribution of $\theta$
494: shows a pattern roughly in accord with the diffusive transport limit,
495: but it also indicates a bimodality divided at $D\sim 15$ Mpc
496: (Fig. 5, {\it top and middle left}).
497: The events with $D \la 15$ Mpc are likely to be cases in which
498: both the source and observer belong to the same filament.
499: These particles are more likely to travel through strongly
500: magnetized filaments rather than void regions, resulting in
501: large deflection angles in addition to a large $\theta_{\rm source}$
502: given by equation (5).
503: On the other hand, for the events with $D \ga 15$ Mpc the particles
504: come from distant sources associated with different filaments.
505: Some of these may fly through voids and arrive with small $\theta$,
506: while most are deflected significantly by the IGMFs.
507: On the other hand, on average the time delay tends to increase with
508: distance $D$, as expected.
509:
510: Super-GZK protons, with $ E_{\rm obs}>60$ EeV come mostly from within $D\sim 100$ Mpc.
511: Most of them from $D \ga 15$ Mpc arrive with $\theta \la 10^{\circ}$
512: and $t_d \la 10^7$ yr (Fig. 5, {\it bottom}).
513: Since the volume filling factor for $B $ greater than a few times $10^{-8}$G
514: (corresponding to $r_g \sim 2-3$ Mpc) is small, those particles could
515: travel almost rectilinearly through void regions, avoiding the
516: strongly magnetized regions of clusters and groups.
517:
518: In Figure 6, the distribution of recorded events is shown in
519: the $({E}_{\rm obs}, \theta)$ and $({E}_{\rm obs},\log t_d$) planes.
520: The events recorded in all six $\Lambda$CDM simulations are included
521: and $\gamma=2.7$ and $D_{\rm min}=3$ Mpc are used.
522: On average the deflection angle decreases with energy, but
523: a clear transition from the diffusive transport regime to the
524: rectilinear propagation regime is apparent around $E_{\rm obs}\sim E_{\rm GZK}$.
525: Sub-GZK particles, with long interaction lengths and small gyroradii,
526: are strongly scattered, while super-GZK particles, with short interaction
527: lengths and large gyroradii, are much less affected.
528: As expected, the time delay decreases with increasing energy
529: on average but has a rather wide spread at any given energy.
530: For comparison, the rectilinear flight time for the mean separation
531: of sources, $l_s=$ 40 Mpc, is $t_{rec}\equiv l_s/c \approx 10^{8}$
532: yr.
533:
534: Figure 7 shows the fractions of recorded events in all six
535: $\Lambda$CDM simulations as functions of $\theta$ and $t_d$,
536: $df/d\theta$ and $df/d \log t_d$, and their cumulative distributions.
537: The events are divided into three energy channels as in Figure 5,
538: and each curve is normalized by the total number of events in
539: the corresponding channel.
540: In order to demonstrate the dispersion of the statistics due to cosmic
541: variance, we also plot the error bars in the cumulative distributions,
542: which are calculated as the standard deviations of the values of $f$
543: for the six simulations.
544:
545: In the lowest energy channel ($10~{\rm EeV}<E_{\rm obs}<30$ EeV, {\it red lines}),
546: the deflection angle is quite large, with about 70\% of the events arriving
547: with $\theta > 30^{\circ}$, that is, $f(>30^{\circ}) \approx 0.7$.
548: Moreover, with $f(>10^8\ {\rm yr}) \approx 0.7$ the time delay
549: is much longer than the typical lifetimes of AGNs ($\tau_{AGN} = 0.01-0.1$
550: Gyr).
551: In the highest energy channel ($E_{\rm obs} \ge 60$ EeV, {\it black lines}),
552: on the other hand, about 70 \% of the recorded events arrive with a
553: deflection angle smaller than $\sim 15^{\circ}$ and a time delay
554: less than $\sim 10^7$ yr.
555: About 35 \% arrive with an angle smaller than $\sim 5^{\circ}$.
556: This implies that the arrival direction of super-GZK cosmic rays may show
557: a positional correlation with the source AGNs and also with the LSS,
558: and the source AGNs are very likely to still be active for such events.
559: We note, however, that these results are not restricted to the specific
560: AGN model.
561: They can be applied for any UHECR sources that have
562: a spatial distribution and magnetic field environment similar
563: to those of luminous X-ray clusters.
564:
565: We note that the present work produces results different
566: from what previous studies have predicted \citep{armen05,dolagetal05}.
567: Specifically, the deflection angle is smaller than that found by
568: Sigl \etal but larger than that of Dolag \etal.
569: This should be attributable to the difference in the models for the IGMFs,
570: as discussed in \S 2.1.
571: We also note that the effects of Galactic magnetic fields are
572: not included in our analysis.
573: Recently, \citet{takami07}, for instance, considered several
574: different models for the Galactic magnetic fields and predicted
575: that the deflection angle of $10^{19.8}$ eV protons should be greater
576: than 8$^{\circ}$ toward the Galactic center, while being mostly $3-5^{\circ}$
577: outside a circular region of 30$^{\circ}$ radius around the
578: Galactic center.
579:
580: \subsection{Predicted Energy Spectrum}
581:
582: Here we present the energy flux, $J(E_{\rm obs})$, of the recorded
583: cosmic-ray events in our propagation experiment.
584: By applying a weighting factor proportional to $E_{\rm inj}^{-\gamma}$
585: to each recorded particle, the energy spectra for different values of
586: $\gamma$ can be constructed.
587: The predicted spectra are calculated for injection spectra
588: with $\gamma$-values of 2.0, 2.4, and 2.7 and for $D_{\rm min}=3$ and
589: $D_{\rm min}=10$ Mpc.
590: Again, all data from the six $\Lambda$CDM simulations are combined.
591: Figure 8 shows the resulting spectra, $J(E)$, along with the data
592: observed at AGASA \citep{nagano00}, HiRes-I \citep{bgg06}, HiRes-II
593: \citep{zech04}, and Auger \citep{parizot07}.
594: Since the amplitude of the injection spectrum is not specified,
595: the amplitude of the predicted $J(E)$ is arbitrary in our model.
596: Therefore each curve was adjusted by eye to fit the HiRes data below
597: $E_{\rm GZK}$.
598:
599: The presence of GZK suppression above 60 EeV is obvious in all the
600: predicted spectra and the observed data except for the AGASA data.
601: The predicted spectra for $\gamma=2.4-2.7$ are all consistent with
602: the observed data, again with the exception of the AGASA data.
603:
604: As shown in the bottom panel of Figure 8, our work predicts that
605: above $\sim 100$ EeV the flux is much higher with $D_{\rm min}=3$
606: Mpc than with $D_{\rm min}=10$ Mpc, indicating that the contribution
607: from nearby sources is important.
608: Thus, if the injection spectrum has a power-law distribution
609: extending well beyond the GZK energy as we assume here,
610: the implication is that the Auger
611: experiment, which has Centaurus A in its field of view, may see
612: a higher flux of super-GZK cosmic rays, compared with experiments in
613: the Northern Hemisphere such as HiRes and the Telescope Array \citep{fm07}.
614: However, it is quite possible that the injection spectrum is limited
615: to a maximum energy $E_{\rm max}$ set by the age and size
616: of the astrophysical accelerators or by the GZK energy loss at the acceleration
617: sites. Moreover, the value of $E_{\rm max}$ may vary with the properties
618: of the accelerator, rather than maintaining a constant value of $10^3$ EeV.
619: In order to settle this issue,
620: much better statistics for the energy spectrum and
621: the arrival directions above 100 EeV are needed.
622:
623: \section{Summary and Discussion}
624:
625: In the search for the astrophysical sources of UHECRs,
626: it is important to understand how the propagation of these
627: charged particles is affected by intergalactic magnetic fields
628: in the large-scale structure of the universe.
629: On the other hand, the information imprinted on
630: the distribution of the UHECR arrival directions may help us
631: to understand the nature of the IGMFs and their roles in
632: the formation and evolution of the LSS and constituent galaxies.
633: Considering the limitations of current observational techniques in
634: measuring the IGMFs in very low density regions such as filaments and
635: voids, it is crucial to construct a physically motivated model
636: to estimate the IGMFs in the LSS.
637:
638: In this study, we adopted a new model based on the turbulence dynamo
639: \citep{ryuetal08} to predict the strength of the IGMFs.
640: The magnetic field energy is estimated from the local vorticity and
641: turbulent kinetic energy of flow motions in cosmological simulations
642: of the LSS formation in a concordance $\Lambda$CDM universe.
643: For the direction of the IGMFs, the topology of the passive magnetic
644: fields followed in the cosmological simulations is used.
645: This approach provides an IGMF model that is independent
646: of the initial seed fields and does not require any renormalization
647: to yield the observed field strength in the intracluster medium.
648: We predict highly structured IGMFs with characteristic field
649: strengths on the order of $10^{-6}$ G in clusters of galaxies and
650: $10^{-8}$ G in filaments.
651: The fields should be much weaker in sheets and voids.
652:
653: Protons with 10 EeV $\le E_{\rm inj}\le 10^3$ EeV are injected at the
654: locations of luminous X-ray clusters with $kT>1$ keV.
655: These sources may represent a population of AGNs
656: residing inside host clusters.
657: This X-ray temperature criterion naturally places the sources at strongly
658: magnetized regions with $ B_s \sim 0.1 \mu$G with a comoving density
659: of $2-3 \times 10^{-5}(\Mpc)^{-3}$.
660: Then the propagation of the UHE protons is followed through the
661: structured IGMFs, including the energy losses due to interactions
662: with the cosmic background radiation
663: The UHE protons are recorded at the positions of mock observers located in groups
664: of galaxies, with halo temperatures in the range
665: $0.05 {\rm keV} < kT < 0.5 {\rm keV}$.
666:
667: Below the GZK energy, the UHE protons come from sources as distant as
668: $\sim 1$ Gpc.
669: They are significantly scattered by the IGMFs, resulting in a wide
670: range of deflection angles, up to 180$^{\circ}$, and have time
671: delays ranging from $10^7$ to $10^{9.5}$ yr.
672: On the other hand, the protons above 60 EeV come mostly
673: from sources within $\sim 100$ Mpc.
674: About 70\% of them avoid strong deflection and arrive at the observers
675: within $\sim 15^{\circ}$ of their source position on the sky with
676: a time delay of less than $\sim 10^7$ yr.
677: About 35\% arrive within $\sim 5^{\circ}$.
678: This implies that there may exist a correlation between the arrival
679: direction of super-GZK cosmic rays and the sky position of the corresponding AGNs.
680: We thus conclude that in the present scenario, UHECR astronomy may be
681: possible at $E>$ 60 EeV.
682: Our prediction seems to be consistent with a recent report by the
683: Auger Collaboration \citep{augerScience} in which the
684: arrival directions of cosmic rays above 60 EeV in their data were found to be
685: correlated with the sky position of AGNs within 75 Mpc.
686:
687: For any cosmological sources, we expect to see GZK suppression
688: in the energy spectrum of UHECRs if the injection spectrum
689: has a power-law distribution and extends well beyond the
690: GZK energy.
691: In this case, nearby sources, within $10 - 20$ Mpc, are expected to
692: make a significant contribution to the flux above $\sim 100$ EeV.
693:
694: Finally, as recently reported by Auger \citep{unger07}, some
695: UHECRs might be heavy nuclei.
696: In the cosmological-shock model, for example, protons can be
697: accelerated up to a few times 10 EeV, so heavy nuclei should
698: dominate the particle flux above that energy
699: \citep{krb97,inoue07}.
700: In a future study, we will consider the propagation of heavy nuclei
701: from cosmological sources in our model IGMFs, taking into account
702: photo disintegration, photo pair production, and photo pion production
703: processes.
704: We expect that the propagation of UHE iron nuclei ($Z=26$), at least, will
705: be in the diffusive transport regime as a results of their much smaller gyroradius
706: ($r_g\propto E/Z$) \citep{armen05}.
707: For intermediate-mass nuclei such as He, C, N, and O, detailed propagation
708: simulations including secondary particles produced by
709: photo disintegrations are necessary in order to determine whether astronomy with
710: UHE nuclei is possible or not.
711:
712: \acknowledgements
713: The work of H. K. and S. D. is supported by Korea Science and Engineering Foundation
714: through the Astrophysical
715: Research Center for the Structure and Evolution of the Cosmos (ARCSEC).
716: The work of D. R. and J. C. is supported by Korea Research Foundation grants
717: funded by the Korean Government (MOEHRD) (KRF-2007-341-C00020 and
718: KRF-2006-331-C00136, respectively).
719: This work was also supported by the Korea Foundation for International Cooperation of
720: Science and Technology through
721: the grant K20702020016-07E0200-01610.
722:
723: \begin{thebibliography}{}
724:
725: \bibitem[Abbasi \etal(2005)]{abbasi05}
726: Abbasi, R. U. \etal (High Resolution Fly's Eye Collaboration)
727: 2005, \apj, 622, 910
728: % composition with HiRes
729: % composition is protons
730:
731: \bibitem[Abbasi \etal(2006)]{abbasi06}
732: Abbasi, R. U. \etal (High Resolution Fly's Eye Collaboration)
733: 2006, \apj, 636, 680
734: % No correlations with BL Lac (HiRes)
735:
736: \bibitem[Armengaud \etal(2005)]{armen05}
737: Armengaud, E., Sigl, G., \& Miniati, F. 2005, Phy. Rev. D., 72, 043009
738:
739: \bibitem[Armengaud \etal(2007)]{armen07}
740: Armengaud, E. \etal, (Pierre Auger Collaboration) 2007,
741: Proc. 30th Int. Cosmic-Ray Conf. (Merida), arXiv0706.2640
742: %Auger:isotropy
743:
744: \bibitem[Berezinsky \& Grigor'eva(1988)]{berezinsky88}
745: Berezinsky, V. S., \& Grigor'eva, S. I. 1988, \aap, 199, 1
746:
747: \bibitem[Berezinsky \etal(2006)]{bgg06}
748: Berezinsky, V. S., Gazizov, A., \& Grigor'eva, S. I.
749: 2006, Phys. Rev. D., 74, 043005
750:
751: \bibitem[Biermann \& Strittmatter(1987)]{biermann87}
752: Biermann, P. L. \& Strittmatter, P. A. 1987, \apj, 322, 643
753: %AGN jets
754:
755: \bibitem[Blasi \& De Marco(2004)]{bm04}
756: Blasi, P. \& De Marco, D. 2004, Astropart. Phys., 20, 559
757:
758: % distant & weak sources
759: \bibitem[Burgett \& O'Malley(2003)]{bo03}
760: Burgett, W. S. \& O'Malley, M. R. 2003, Phys. Rev. D., 67, 092002
761:
762: % MF affects particle propagation
763: \bibitem[Dolag \etal(2004)]{dolagetal04}
764: Dolag, K., Grasso, D., Springel, V., \& Tkachev. I. 2004,
765: JETP. Lett., 79, 583
766:
767: \bibitem[Dolag \etal(2005)]{dolagetal05}
768: Dolag, K., Grasso, D., Springel, D., \& Tkachev, J.
769: 2005, Journal of Cosmology and Astroparticle Physics, 1, 009
770:
771: \bibitem[Facal \etal(2007)]{facal07}
772: Facal, P. \etal, (Pierre Auger Collaboration) 2007,
773: Proc. 30th Int. Cosmic-Ray Conf. (Merida), arXiv0706.4322
774: %Auger energy spectrum
775:
776: \bibitem[Ferrarese \etal(2007)]{ferrarese07}
777: Ferrarese, L, Mould, J. R., Stetson, P. B., Tonry, J. L.,
778: Blakeslee, J. P., \& Ajhar, E. S. 2007, \apj, 654, 186
779: % Cepheids and distance to Centaurus A
780:
781: \bibitem[Fukushima \etal(2007)]{fm07}
782: Fukushima, M. et al. (Telescope Array collaboration) 2007,
783: Proc. 30th Int. Cosmic-Ray Conf. (Merida)
784:
785: \bibitem[Harari \etal(2007)]{harari07}
786: Harari, D. \etal, (Pierre Auger Collaboration) 2007,
787: Proc. 30th Int. Cosmic-Ray Conf. (Merida), arXiv0706.1715
788: %Auger BL Lac
789:
790: \bibitem[Hillas(1984)]{hillas84}
791: Hillas, A. M. 1984, \araa, 22, 425
792: %The Origin of UHECRs
793:
794: \bibitem[Inoue \etal(2007)]{inoue07}
795: Inoue, S., Sigl, G., Miniati, F., \& Armengaud, E. 2007,
796: preprint (astro-ph/0701167)
797:
798: % GZK cutoff
799: \bibitem[Greisen (1966)]{g66}
800: Greisen, K. 1966, Phys. Rev. Lett., 16, 748
801:
802: \bibitem[Kang \etal(1997)]{krb97}
803: Kang, H., Rachen, J. P., \& Biermann, P. L. 1997, \mnras, 286, 25
804: % cluster accretion shock
805:
806: \bibitem[Kang \etal(1996)]{krj96}
807: Kang, H., Ryu, D., \& Jones, T. 1996, \apj, 456, 422
808: % cluster accretion shock sources
809:
810: \bibitem[Kang \etal(2007)]{krco07}
811: Kang, H., Ryu, D., Cen, R., \& Ostriker, J. P. 2007, \apj, 669, 729
812: % cluster accretion shock sources
813:
814: \bibitem[Kulsrud \etal(1997)] {kulsrude97}
815: Kulsrud, R. M., Cen, R., Ostriker, J. P., \& Ryu, D.
816: 1997, \apj, 480, 481
817: % Biermann battery mechanism
818:
819: \bibitem[Mei \etal(2007)]{mei07}
820: Mei, S., Blakeslee, J. P., Cote, P., Tonry, J. L., West, M. J.,
821: Ferrarese, L., Jordan, A., Peng, E. W., Anthony, A., \& Merritt, D. 2007,
822: \apj, 655, 144
823:
824: \bibitem[Nagano \& Watson (2000)]{nagano00}
825: Nagano, A. \& Watson, A. A.,
826: 2000, Rev. Mod. Phys., 72, 689
827:
828: \bibitem[Parizot \etal(2007)]{parizot07}
829: Parizot, E. \etal, (Pierre Auger Collaboration) 2007,
830: proceedings for the the International Symposium on Astronomy and Astrophysics
831: of Extreme Universe, arXiv0709.2500
832: %The Pierre Auger Observatory: status, results and perspective
833:
834: \bibitem[Abraham \etal(2007)]{augerScience}
835: Abraham, J. et al. 2007, Science, 318, 939
836: % Recent Auger Result
837:
838: \bibitem[Pravdin \etal(1999)]{pravdin99}
839: Pravdin, M. I. et al. (Yakutsk collaboration) 1999,
840: proceedings for the 26th ICRC (Salt Lake City, USA), 3, 292
841: %Yakutsk data:J(E)
842:
843: % Local Group temperature
844: \bibitem[Rasmussen \& Pedersen (2001)]{rasmus01}
845: Rasmussen, J. \& Pedersen, K. 2001, \apj, 559, 892
846:
847: \bibitem[Ryu \etal(1998)]{rkb98}
848: Ryu, D., Kang, H., \& Biermann, P. L. 1998, \aap, 335, 19
849:
850: \bibitem[Ryu \etal(2008)]{ryuetal08}
851: Ryu, D., Kang, H., Cho, J., \& Das, S. 2008, Science, 320, 909
852:
853: \bibitem[Sigl \etal(2003)]{sigletal03}
854: Sigl, G., Miniati, F., \& Ensslin, T. A. 2003,
855: Phys. Rev. D., 68, 043002
856:
857: \bibitem[Sigl \etal(2004)]{sigletal04}
858: Sigl, G., Miniati, F., \& Ensslin, T. A. 2004,
859: Phys. Rev. D., 70, 043007
860: % calculation of deflection angle
861:
862: \bibitem[Takami \& Sato(2007)]{takami07}
863: Takami, H. \& Sato, K. 2007, arXiv0711.2386
864: %GMF effects
865:
866: \bibitem[Takeda \etal(1999)]{takeda99}
867: Takeda, M. \etal 1999, ApJ, 522, 225
868: %small-scale clustering of AGASA
869:
870: \bibitem[Tinyakov \& Tkachev(2001)]{tinyakov01}
871: Tinyakov, P. G. \& Tkachev, I. I. 2001, JETP Lett., 74, 445
872: % AGASA-BL Lac
873:
874: \bibitem[Unger \etal(2007)]{unger07}
875: Unger, M. \etal, (Pierre Auger Collaboration) 2007,
876: Proc. 30th Int. Cosmic-Ray Conf. (Merida), arXiv0706.1495
877: %Auger X_max
878:
879: \bibitem[Uchihori \etal(2000)]{uchihori00}
880: Uchihori, Y., Nagano, M., Takeda, M., Teshima, M.,
881: Lioyd-Evans, J., \& Watson, A. A. 2000, Astropart. Phys., 13, 151
882: % Multiplets from same direction
883:
884: \bibitem[Watson(2006)]{watson06}
885: Watson, A. A. 2006, Nucl. Phys. Proc. Suppl., 151, 83
886: % Chemical composition, hadron instead of photon
887:
888: \bibitem[Waxman(1995)]{waxman95}
889: Waxman, E. \prl, 75, 386
890:
891: % AGN sources
892: \bibitem[Yoshiguchi \etal(2003)]{Yoshiguchietal03}
893: Yoshiguchi, H., Nagataki, S., \& Sato, K. 2003, \apj, 592, 311
894:
895: \bibitem[Zatsepin \& Kuzmin (1966)]{zk66}
896: Zatsepin, G. T. \& Kuzmin, V. A. 1966, JETP Lett., 4, 78
897:
898: \bibitem[Zech (2004)]{zech04}
899: Zech, A. \etal (High Resolution Fly's Eye Collaboration)
900: 2004, Nuclear Physics B (Proc. Suppl.), 136, 34
901:
902: \end{thebibliography}
903:
904:
905:
906: \clearpage
907:
908: % ----------- Table I ------------------------------------------------
909: \begin{deluxetable}{cccc}
910: %\begin{deluxetable}{rrrrr}
911: \tablecolumns{3}
912: \tablewidth{0pc}
913: \tablecaption{Numbers of Sources and Observers}
914: \tablehead{
915: \colhead{Simulation} & \colhead{Sources} & \colhead{Observers}}
916: \startdata
917: $\Lambda$CDM1 & 18 & 1000 \\
918: $\Lambda$CDM2 & 31 & 1344 \\
919: $\Lambda$CDM3 & 20 & 1379 \\
920: $\Lambda$CDM4 & 29 & 1336 \\
921: $\Lambda$CDM5 & 24 & 1343 \\
922: $\Lambda$CDM6 & 28 & 1365 \\
923: %
924: \enddata
925: \end{deluxetable}
926:
927: %---------- figure 1 ---------------------------------------------
928: \clearpage
929: \begin{figure}
930: \vspace{-12cm}
931: \centerline{\includegraphics[width=18cm]{f1.eps}}
932: \vspace{-4cm}
933: \figcaption{ {\it Left}:
934: Two-dimensional slice of area $(100 \Mpc)^2$
935: showing the distribution of magnetic field
936: strength in our model at redshift $z=0$.
937: {\it Right}:
938: Blown-up image of the box delineated with dotted lines in the left panel.
939: The contour levels are color-coded as follows:
940: $\log B=$ $-12$ (red), $-11$ (blue), $-10$ (cyan), $-9$ (black), $-8$
941: (magenta) and $-7$ (green), where $B$ is the field strength in units
942: of gauss.
943: \label{fig1}}
944: \end{figure}
945:
946: %---------- figure 2 ---------------------------------------------
947: \clearpage
948: \begin{figure}
949: \vspace{-13cm}
950: \centerline{\includegraphics[width=18cm]{f2.eps}}
951: \vspace{-4cm}
952: \figcaption{
953: {\it Left}: volume fraction in the gas density vs. field strength plane with
954: our model IGMFs at redshift $z = 0$.
955: {\it Right}: volume fraction, $df/d \log B $ ({\it solid line}),
956: and its cumulative distributions, $f(>B)$ ({\it dotted line}) and
957: $f(<B)$ ({\it dot-dashed line}) as a function of the IGMF strength.
958: \label{fig2}}
959: \end{figure}
960:
961: %---------- figure 3 ---------------------------------------------
962: \clearpage
963: \begin{figure}
964: \vspace{-1cm}
965: \centerline{\includegraphics[width=12cm]{f3.eps}}
966: \figcaption{
967: Distribution of UHECR sources ({\it red circles}) and mock observers ({\it blue circles})
968: in simulation $\Lambda$CDM1.
969: Sources are modeled as AGNs inside X-ray clusters, while mock observers
970: are placed inside groups of galaxies similar to the Local Group.
971: \label{fig3}}
972: \end{figure}
973:
974: %---------- figure 4 ---------------------------------------------
975: \clearpage
976: \begin{figure}
977: \vspace{-12cm}
978: \centerline{\includegraphics[width=18cm]{f4.eps}}
979: \vspace{-4cm}
980: \figcaption{
981: {\it Left}: Distribution of magnetic field strength at source
982: locations in the six simulations with different initial conditions.
983: Different line styles are used for simulations $\Lambda$CDM1 - $\Lambda$CDM6.
984: {\it Right}:
985: Distribution of magnetic field strength within the observer spheres.
986: The same line styles as in the left panel are used.
987: \label{fig4}}
988: \end{figure}
989:
990: %---------- figure 5 ---------------------------------------------
991: \clearpage
992: \begin{figure}
993: \vspace{-1cm}
994: \centerline{\includegraphics[width=18cm]{f5.eps}}
995: \figcaption{
996: Deflection angle ($\theta$) and time delay ($t_d$) as functions of
997: the source-to-observer distance $(D)$.
998: The events recorded with observed energies $ 10~{\rm EeV}\le E_{\rm obs} < 30$
999: EeV are shown in the top panels, $ 30~{\rm EeV} \le E_{\rm obs} < 60$ EeV in the
1000: middle panels, and $E_{\rm obs} \ge 60$ EeV in the bottom panels.
1001: The data points are color-coded by the injection energy as follows:
1002: red, $ 10~{\rm EeV} \le E_{\rm inj} < 30$ EeV; blue, $ 30~{\rm EeV}
1003: \le E_{\rm inj} < 60$ EeV; green, $E_{\rm inj} \ge 60$ EeV.
1004: An injection spectrum $N(E_{\rm inj}) \propto E_{\rm inj}^{-2.7}$
1005: is assumed.
1006: (See text for details.)
1007: \label{fig5}}
1008: \end{figure}
1009:
1010: %---------- figure 6 ---------------------------------------------
1011: \clearpage
1012: \begin{figure}
1013: \vspace{-12cm}
1014: \centerline{\includegraphics[width=18cm]{f6.eps}}
1015: \vspace{-4cm}
1016: \figcaption{
1017: Distribution of observed UHECR events in the planes of observed energy
1018: vs. deflection angle ({\it left}) and time delay ({\it right}).
1019: The events recorded in all six $\Lambda$CDM simulations are included,
1020: and $\gamma=2.7$ and $D_{\rm min}=3$ Mpc were used.
1021: \label{fig6}}
1022: \end{figure}
1023:
1024: %---------- figure 7 ---------------------------------------------
1025: \clearpage
1026: \begin{figure}
1027: \vspace{-1cm}
1028: \centerline{\includegraphics[width=18cm]{f7.eps}}
1029: \figcaption{
1030: Fractions of observed UHECR events as a function of deflection angle
1031: ({\it top left}) and time delay ({\it top right}), and the respective
1032: cumulative distributions ({\it bottom}).
1033: The events recorded in all six $\Lambda$CDM simulations are included,
1034: and $\gamma=2.7$ and $D_{\rm min}=3$ Mpc were used.
1035: The distributions for different energy channels are shown:
1036: red, $10~{\rm EeV} < E_{\rm obs}< 30$ EeV;
1037: blue, $30~{\rm EeV} \le E_{\rm obs} < 60$ EeV;
1038: black, $E_{\rm obs}\ge 60$ EeV.
1039: The error bars shown for the cumulative distributions are the
1040: standard deviations of $f$ for the six simulations.
1041: \label{fig7}}
1042: \end{figure}
1043:
1044: % ----------- figure 8 --------------------------------------------
1045: \clearpage
1046: \begin{figure}
1047: \vspace{-1cm}
1048: \centerline{\includegraphics[width=18cm]{f8.eps}}
1049: \figcaption{
1050: Energy spectra of UHE protons predicted by our model.
1051: The injection spectrum at the sources is proportional to
1052: $E_{\rm inj}^{-\gamma}$ for $10 \le E_{\rm inj} \le 10^3$ EeV.
1053: {\it Top}: The blue, red, and black lines are for
1054: $\gamma$-values of 2.0, 2.4, and 2.7, respectively, with a minimum
1055: source-to-observer distance $D_{\rm min}=3$ Mpc.
1056: {\it Bottom}: The red lines are for an injection spectrum
1057: with $\gamma=2.4$, while the blue lines are for $\gamma=2.7$.
1058: The solid lines are for $D_{\rm min}=3$ Mpc, and the dotted lines
1059: are for $D_{\rm min}=10$ Mpc.
1060: The data observed at AGASA \citep{nagano00}, HiRes-I \citep{bgg06},
1061: HiRes-II \citep{zech04}, and Auger \citep{parizot07} are marked
1062: with asterisks, open circles, filled circles, and triangles,
1063: respectively.
1064: The predicted spectra were arbitrarily scaled by eye to fit the
1065: HiRes data below $E_{GZK}$.
1066: \label{fig8}}
1067: \end{figure}
1068:
1069: \end{document}
1070: