astro-ph0204360/ms.tex
1: \documentclass[12pt,preprint]{aastex}
2: %\documentclass[12pt,manuscript]{aastex}
3: %\renewcommand{\thefootnote}{\fnsymbol{footnote}}
4: %\usepackage[active]{srcltx}
5: \slugcomment{UNIL-IPT-02-1}
6: \shorttitle{$\gamma$-ray versus UHECR emission by BL Lac objects}
7: \shortauthors{Gorbunov et.al.}
8: \begin{document}
9: \title{%
10: Evidence for a connection between the $\gamma$-ray and the highest
11: energy cosmic-ray emissions by BL Lacertae objects
12: }
13: \author{D.~S.~Gorbunov\altaffilmark{1,4},
14: P.~G.~Tinyakov\altaffilmark{1,2,5},
15: I.~I.~Tkachev\altaffilmark{1,3,6},
16: and
17: S.~V.~Troitsky\altaffilmark{1,7}
18: \altaffiltext{1}{Institute for Nuclear Research of the Russian Academy of
19: Sciences,
20: 60th October Anniversary Prospect 7a, 117312, Moscow, Russia}
21: \altaffiltext{2}{Institute of Theoretical Physics, University of Lausanne,
22: CH-1015, Lausanne, Switzerland}
23: \altaffiltext{3}{CERN Theory Division, CH-1211 Geneva 23, Switzerland}
24: \altaffiltext{4}{e-mail: {\tt gorby@ms2.inr.ac.ru}}
25: \altaffiltext{5}{e-mail: {\tt Peter.Tinyakov@cern.ch}}
26: \altaffiltext{6}{e-mail: {\tt Igor.Tkachev@cern.ch}} 
27: \altaffiltext{7}{e-mail: {\tt st@ms2.inr.ac.ru}}
28: }
29: \date{}
30: \begin{abstract}
31: A set of potentially $\gamma$-ray--loud BL Lac objects is selected by
32: intersecting the EGRET and BL Lac catalogs. Of the resulting 14
33: objects, eight are found to correlate with arrival directions of
34: ultra-high-energy cosmic rays (UHECRs), with significance of the order
35: of $5\sigma$.  This suggests that $\gamma$-ray emission can be used as
36: a distinctive feature of those BL Lac objects that are capable of
37: producing UHECR.
38: \end{abstract}
39: \keywords{cosmic rays --- BL Lacertae objects: general 
40: --- gamma rays: theory}
41: 
42: 
43: The highest energy cosmic rays with energies in excess of $10^{19}$~eV
44: (ultra-high-energy cosmic rays [UHECRs]), observed by AGASA \citep{AG}
45: and Yakutsk \citep{YK} experiments, show a significant number of
46: clusters at angles of the order of experimental angular resolution
47: \citep{clusters1}. The significance of clustering is quantitatively
48: estimated by calculating the angular correlation function of the UHECR
49: events \citep{Tinyakov:2001ic}. It follows that the observed
50: clustering has probability of less than $10^{-5}$ to occur as a result
51: of a statistical fluctuation. This suggests that (1) there exist
52: compact sources of UHECRs and (2) the already existing data may
53: contain information sufficient to identify the actual sources, the
54: subset of cosmic rays with maximum autocorrelations being the best
55: choice for this purpose.
56: 
57: This line of reasoning was pursued by \citet{Tinyakov:2001nr} assuming
58: that BL Lacertae objects are relevant candidates.  Significant
59: correlations were found with the subset of most powerful confirmed BL
60: Lac objects. After assigning penalties for subset selection and bin
61: size adjustment, the probability of such correlation to occur by
62: chance in a random distribution is of order $10^{-4}$. BL Lac objects
63: comprise a subclass of blazars characterized by the absence of
64: emission lines. Blazars are thought to have relativistic jets directed
65: along the line of sight, while the absence of emission lines indicates
66: low ambient matter and radiation fields and therefore favorable
67: conditions for the acceleration of particles to highest energies. For
68: this reason, BL Lac objects may be particularly promising candidates
69: for UHECR sources.
70: 
71: It follows from both the statistical arguments \citep{Dubovsky:2000gv}
72: and correlation analysis \citep{Tinyakov:2001nr,Tinyakov:2001ir} that
73: only a small fraction of existing BL Lac objects should be capable of
74: producing highest energy cosmic rays. For understanding the nature of
75: the sources and the mechanism of UHECR emission, the key question is
76: which physical characteristics single out the actual emitters among
77: all BL Lac objects? In this Letter we propose that the strong
78: $\gamma$-ray emission is the feature that distinguishes UHECR sources.
79: 
80: There are general reasons to expect the connection between UHECR and
81: $\gamma$-ray emissions. Both the acceleration of particles in the
82: source and their subsequent propagation in the intergalactic space is
83: accompanied by energy losses. A substantial part of this energy is
84: transferred into the electromagnetic cascade and, generically, ends up
85: in the EGRET energy region \citep{Berezinsky,Aharonian}.  In models
86: involving neutrinos via the Z-burst mechanism
87: \citep{Fargion,Weiler,Fargion:2001pu}, and those based on very
88: high-energy photons \citep{Kalashev:2001qp,Neronov:2002se}, the
89: astrophysical accelerator must be very powerful to provide sufficient
90: flux of primary ultra--high-energy particles. In these models, one may
91: expect a strong electromagnetic radiation from the source and
92: substantial contribution into EGRET flux. Note that the {\em
93: extragalactic} cascade may get isotropized by random magnetic fields
94: when approaching the low energy end; this may cause smearing of point
95: sources and result in contributions into $\gamma$-ray background. In any
96: case, these arguments suggest that $\gamma$-ray emission may be an
97: important distinctive feature of UHECR sources\footnote{We are
98: grateful to A. Neronov and D. Semikoz for numerous useful discussions
99: of this subject.  Note that possible connection between gamma and
100: neutrino signals has also been discussed in \citet{Fargion2}.}.
101: 
102: In order to test this hypothesis we first select those BL Lac objects
103: that can be associated with $\gamma$-ray sources and then study their
104: correlations with UHECR. The most complete list of the $\gamma$-ray
105: sources can be found in the third EGRET catalog \citep{3EG} containing
106: 271 object. Of these objects, 67 are identified with active galactic
107: nuclei (AGNs), five with pulsars, one with a solar flare, one with the
108: LMC, and 27 are tentatively identified with AGNs. The remaining 170
109: objects are unidentified.
110: 
111: In this Letter we do not rely on the existing EGRET identification of
112: objects, neither do we attempt our own object-by-object analysis.
113: Instead, we adopt a purely statistical approach: we take the full set
114: of confirmed BL Lac objects from the Veron2001 catalog \citep{Veron}
115: consisting of 350 objects, and we select a subsample of those that may
116: be associated with an EGRET $\gamma$-ray source. The selection
117: procedure is as follows: Point sources in the EGRET catalog are
118: defined as a local excess of a signal over the background. Each source
119: is associated with a contour containing 95\% of the signal. For each
120: contour, a circle of equal area is defined, with the radius $R_{95}$.
121: These radii are listed in the EGRET catalog. They roughly correspond
122: to uncertainties in the positions of the sources.  However, the 95\%
123: contours are often noncircular. Additional systematic errors in
124: position determination may be present in the case of a bright nearby
125: source (such cases are marked as ``confused'' in the catalog). As a
126: result, many well-identified sources (e.g., the Vela pulsar that is
127: unambiguously identified by timing) fall outside of $R_{95}$. In our
128: analysis, we consider an object to be associated with the EGRET source
129: if the angular distance between the two does not exceed $2R_{95}$. In
130: cases of ambiguity the nearest neighbor is taken.
131: 
132: According to this procedure, 14 BL Lac objects from the Veron2001
133: catalog are associated with EGRET sources. They are listed in Table~1.
134: Of these 14 objects, eight already have identifications in the EGRET
135: catalog, while six are newly proposed identifications. Out of eight
136: previously identified objects, five have the same identifications in
137: the SIMBAD database as is suggested by our procedure (objects marked
138: by asterisks in Table~1). Interestingly, in those three cases when our
139: procedure suggests identification different from the existing one, the
140: latter has a question mark in the SIMBAD database, while in five cases
141: when they coincide the existing identification is considered
142: firm. This rather good agreement with previous results gives
143: confidence that at least part of previously unidentified EGRET sources
144: listed in Table~1 should be identified with corresponding BL Lac
145: objects.
146: 
147: Since the EGRET 95\% contours are large enough to contain several
148: astrophysical objects, the identifications depend on the assumptions
149: about candidate sources.  Most previous works have concentrated on the
150: powerful radio quasars as possible candidates (see, e.g.,
151: \citet{identifications}). An approach somewhat similar to ours was
152: used by \citet{punsly} where correlations of EGRET catalog with X-ray
153: and moderate radio sources ({\it ROSAT}--Green Bank catalog) were
154: considered. It revealed several new identifications, large fraction of
155: them being BL Lac objects.
156: 
157: Being based on position coincidence only, the identifications proposed
158: in Table~1 cannot be considered as final. Instead, Table~1 should be
159: treated as a starting point for more detailed object-by-object study,
160: including EGRET intensity maps, time correlations, etc. Such an
161: analysis goes beyond the scope of this Letter. It is important to
162: note, however, that possible misidentifications in Table~1 do not
163: compromise our main result, strong correlation of the selected
164: subsample with UHECRs. Like any random factor, such misidentifications
165: can only diminish the correlations.
166: 
167: Let us now turn to correlations between the set of 14 (potentially)
168: $\gamma$-ray--loud BL Lac objects of Table~1 and UHECRs. In the part
169: concerning UHECRs, we follow the approach of \citet{Tinyakov:2001nr}
170: and use the set of cosmic rays with the largest autocorrelations.
171: This set consists of 39 AGASA events with energies $E>4.8\times
172: 10^{19}$~eV and 26 Yakutsk events with energies $E>2.4\times
173: 10^{19}$~eV \citep{Tinyakov:2001ic}.
174: 
175: The numerical algorithm used in this Letter is identical to that of
176: \citet{Tinyakov:2001ic,Tinyakov:2001nr,Tinyakov:2001ir}. We
177: characterize the significance of correlations between UHECRs and a
178: given set of sources at a given angular scale $\delta$ by the
179: probability $p(\delta)$ defined in the following way. First, we count
180: the number of source/cosmic-ray pairs separated by the angle $\leq
181: \delta$ in the real data, thus obtaining the data count
182: $N_d(\delta)$. We then generate a large number of random (mock) sets
183: of cosmic rays, taking into account actual acceptance of the
184: experiments in such a way that the large-scale distribution of mock
185: cosmic rays is uniform. On small scales we introduce autocorrelations
186: in mock sets since the real data are clustered. The number of clusters
187: added in each mock set mimics the real data, while cluster positions
188: are random. For each mock set, the number of source/cosmic-ray pairs
189: is then counted in the same way as for the real data, giving the mock
190: count $N_m(\delta)$. At a large total number of mock sets, the
191: fraction of mock sets for which $N_m(\delta)\geq N_d(\delta)$ gives
192: $p(\delta)$.
193: 
194: In the correlation analysis, we take into account possible effects of
195: the Galactic magnetic field (GMF) on propagation of UHECRs. We use the
196: spiral model of GMF with different directions of the field in the two
197: spiral arms and consider two cases: symmetric and antisymmetric
198: field with respect to the galactic plane. The details of the model and
199: corresponding parameters can be found in \citet{Tinyakov:2001ir}
200: together with further references. We assume that primary particles can
201: have charges of $Q=0,\pm 1$. 
202: 
203: In the case $Q\neq 0$, the positions of cosmic rays are corrected for
204: the deflections in GMF prior to counting the number of pairs with
205: given angular separation. For each cosmic ray there are several
206: possible positions after correction for GMF corresponding to different
207: allowed charges. For a given ray, the minimum angular distance over
208: the set of sources and charges determines the resulting charge
209: assigned to that ray. In all cases, each randomly generated set is
210: subject to exactly the same procedure as the real data. This
211: guarantees that no correlations are artificially introduced.
212: 
213: The results of the calculations for the charge assignments $Q=0$,
214: $Q=1$, $Q=0,1$ and $Q=0,\pm1$ and for two types of magnetic field
215: (symmetric and antisymmetric) are presented in Table~2. The
216: correlations are rather significant in all cases, being the best in
217: the case of charges $Q=0,1$ and antisymmetric field, in agreement with
218: \citet{Tinyakov:2001ir}. In this case, the data count
219: $N_d(2.7^{\circ})= 13$, while 2 is expected in average for a uniform
220: background.  The probability for this to occur by chance is
221: $p(\delta)=3\times 10^{-7}$ ($5.1\sigma$). The dependence of
222: $p(\delta)$ on $\delta$ in this case is shown in Fig.~\ref{SS-EV}.
223: % 
224: \if 0
225: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
226: \begin{figure}
227: \plotone{f1.eps}
228: \caption{Significance of correlations between 14 $\gamma$-ray loud BL
229: Lac objects and UHECR as a function of the angular scale $\delta$ for
230: the $Q=0,1$ charge composition. This corresponds to the lowest
231: probability entry of the Table 2.}
232: \label{SS-EV}
233: \end{figure}
234: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
235: \fi 
236: 
237: {}From the analysis of Table~2 one is tempted to conclude that the
238: case $Q=0,1$ (neutral and positively charged particles) is favored.
239: However, present statistics are not enough for such a conclusion, as
240: the following simple argument shows. The data count is subject to
241: fluctuations that may be roughly estimated as $\pm\sqrt{N_d}$ (these
242: fluctuations would be observed if the AGASA and Yakutsk experiments
243: were repeated many times). If the ``average'' data count were 10,
244: counts from seven to 13 would occur equally often. Corresponding
245: probabilities $p(\delta)$ would range from $10^{-4}$ to
246: $10^{-7}$. Thus, unlike correlations themselves, the {\em difference}
247: between the cases $Q=0$, $Q=0,1$ and $Q=0,\pm 1$ can easily be
248: explained by a fluctuation.
249: 
250: Energies and charges of UHECR events that contribute into correlations
251: with $\gamma$-ray--loud BL Lac objects are listed in columns (7) and
252: (8) of Table~1 (the antisymmetric magnetic field model is
253: assumed). Multiple charges in column (8) mean that the corresponding
254: event contributes to correlations under different charge assignments.
255: 
256: The comparison between Table~1 of this Letter and Table~1 of
257: \citet{Tinyakov:2001nr} shows that the same BL Lac objects and
258: cosmic rays contribute to correlations in Ref.~\citep{Tinyakov:2001nr}
259: and in the case of $Q=0$ presented above. In \citet{Tinyakov:2001nr},
260: the set of brightest BL Lac objects was selected by imposing cuts on
261: redshift, apparent magnitude, and radio flux. In the resulting subset
262: of 22 BL Lac objects, five candidate sources were identified. It is
263: remarkable that four out of these five candidates, in particular all
264: three that correlate with UHECR multiplets, are among the 14 BL Lac
265: objects that comprise the intersection of BL Lac and EGRET catalogs,
266: $\gamma$-ray--loud BL Lac objects. Even more remarkable is that out
267: of 10 remaining BL Lac objects, four correlate with cosmic rays after
268: correction for GMF. Among the remaining six that do not correlate with
269: UHECRs, two objects are situated in the Southern hemisphere invisible
270: for Yakutsk and AGASA experiments. These objects can be excluded from
271: correlation analysis. Thus, the majority of $\gamma$-ray--loud BL Lac
272: objects (eight out of 12) correlate with UHECR. One concludes that the
273: ability to emit $\gamma$-rays may be used as the physical criterion that
274: allows to select actual UHECR sources from the set of all BL Lac
275: objects.
276: 
277: BL Lac objects are typically faint objects. Some of the unidentified
278: EGRET sources may be actually BL Lac objects that have not not yet
279: been observed at other wavelengths, or have been observed but not
280: identified as BL Lac objects. If this is the case and our conclusion
281: about the connection between $\gamma$-ray and UHECR emissions is
282: correct, one may expect correlations between unidentified EGRET
283: sources and UHECRs. To check this, we calculated correlations between
284: UHECRs and unidentified EGRET sources having Galactic latitude
285: $|b|>10^{\circ}$ (the cut $|b|>10^{\circ}$ is made to increase the
286: fraction of extragalactic sources as, according to \citet{Grenier},
287: the total number of such sources is expected to be 30-40 only). This
288: set contains 96 objects. Correlations are best when all particles are
289: assumed to have a charge of $Q=+1$; corresponding significance is
290: $p(\delta)= 10^{-4}$ at $\delta =3^{\circ}$. Table~3 summarizes EGRET
291: sources and cosmic rays that contribute to correlations. It is
292: interesting to note that, unlike Table~1, Table~3 seems to favor
293: positively charged particles. We expect that some EGRET sources listed
294: in Table~3 are BL Lac objects that have not yet been observed.
295: 
296: To summarize, there exists a significant correlation of arrival
297: directions of UHECRs with $\gamma$-ray--loud BL Lac objects (BL Lac
298: objects that may be associated with the EGRET sources). This confirms
299: the conjecture that strong $\gamma$-ray emission is a characteristic
300: feature of those BL Lac objects that are the sources of UHECR. Present
301: data are compatible with charges of primary particles $Q=0$, $Q=+1$,
302: $Q=0,+1$ and $Q=0,\pm 1$, although they favor the latter two cases. It
303: does not seem possible, with the present statistics, to distinguish
304: between these cases on the basis of correlation analysis, but it
305: should be possible in the future. This question is of particular
306: interest since specific charge composition is a good signature of most
307: of the existing models.  Charge $Q=0,\pm 1$ would speak strongly for
308: neutrino models. Charge $Q=1$ would favor protons accelerated in BL
309: Lac objects (note that energies of most of the $Q=1$ events in Table~1
310: would allow them to reach us from super-GZK distances provided
311: extragalactic magnetic fields are small). The cases $Q=0,1$ and $Q=0$
312: would suggest, in view of the distance to BL Lac objects and presence
313: of neutral particles, the existence of new physics (e.g., exotic
314: neutral particles \citep{Chung,Berezinsky2,Gorbunov} or violation of
315: Lorentz invariance \citep{Coleman,Dubovsky2}).
316: 
317: The results presented here suggest that the sources of UHECRs are
318: high-energy--peaked BL Lac objects located opposite the flat-spectrum
319: radio quasar end of the ``unified blazar
320: sequence''\citep{blazar-sequence}. This does not contradict the
321: conclusions of \citet{Sigl}, who found no correlations between UHECR
322: and {\em identified\/} EGRET blazars. Indeed, most of the latter are
323: high-polarization blazars, and not low-polarization,
324: high-energy--peaked BL Lac objects that, according to our study, are
325: the most probable sources of UHECR.
326: 
327: The work of P.T.\ is supported by the Swiss Science Foundation, grant
328: 21-58947.99. The work of D.G.\ and S.T.\ is supported in part by the
329: program SCOPES of the Swiss National Science Foundation, project
330: No.~7SUPJ062239, by SSLSS grant 00-15-96626, and by RFBR grant
331: 02-02-17398. The work of D.G., I.T., and S.T. is supported in part by
332: INTAS grants YSF 2001/2-142, 99-1065 and YSF 2001/2-129, respectively.
333: This Letter made use of the SIMBAD database, operated at the Centre de
334: Donn\'ees Astronomiques de Strasbourg, Strasbourg, France.
335: 
336: \begin{thebibliography}{99}
337: \bibitem[Afanasiev et\ al.(1996)]{YK} Afanasiev, B.~N., et al.
338: 1996, in Proc. Int. Symp. on Extremely High Energy Cosmic Rays:
339: Astrophysics and Future Observatories, Ed. by Nagano, p.32
340: 
341: \bibitem[Berezinsky et\ al.(1990)]{Berezinsky}
342: Berezinsky, V.~S., Bulanov S.~V., Dogiel, V.~A., Ginzburg, V.~ L., 
343: \& Ptuskin, V.~S. 1990, Astrophysics of Cosmic Rays (Amsterdam: Elsevier)
344:   
345: \bibitem[Berezinsky et\ al.(2002)]{Berezinsky2}
346: Berezinsky, V., Kachelriess, M., \& Ostapchenko, S. 2002,
347: %``Extensive air showers from ultra high energy gluinos,''
348: \prd, {65}, 083004
349: %%CITATION = ASTRO-PH 0109026;%%
350: 
351: \bibitem[Chung et\ al.(1998)]{Chung}
352: Chung, D.~J., Farrar, G.~R., \& Kolb, E.~W. 1998
353: %``Are ultrahigh energy cosmic rays signals of supersymmetry?,''
354: \prd, {57}, 4606
355: %%CITATION = ASTRO-PH 9707036;%%
356: 
357: \bibitem[Coleman \& Glashow (1999)]{Coleman}
358: Coleman, S.~R. \& Glashow, S.~L. 1999,
359: %``High-energy tests of Lorentz invariance,''
360: \prd, {59}, 116008
361: %%CITATION = HEP-PH 9812418;%%
362: 
363: \bibitem[Coppi \& Aharonian(1996)]{Aharonian}
364: Coppi,~P.~S. \& Aharonian,~F.~A. 1996,
365: %``Constraints on the very high energy emissivity of the 
366: %universe from  the diffuse GeV $\gamma$-ray background,''
367: \apj, 487, L9 
368: %%CITATION = ASTRO-PH 9610176;%%
369: 
370: \bibitem[Dubovsky et\ al.(2000)]{Dubovsky:2000gv}
371: Dubovsky, S.~L., Tinyakov, P.~G., \& Tkachev, I.~I. 2000,
372: \prl, 85, 1154 
373: %%CITATION = ASTRO-PH 0001317;%%
374: 
375: \bibitem[Dubovsky \& Tinyakov (2002)]{Dubovsky2}
376: Dubovsky, S.~L. \& Tinyakov, P.~G. 2002,
377: %``Violation of Lorentz invariance and neutral component of UHECR,''
378: Astropart.\ Phys.\  {18}, 89
379: %%CITATION = ASTRO-PH 0106472;%%
380: 
381: \bibitem[Fargion et\ al.(1999)]{Fargion}
382: Fargion, D., Mele, B., \& Salis, A. 1999, 
383: \apj, 517, 725
384: 
385: \bibitem[Fargion et\ al.(2001)]{Fargion:2001pu}
386: Fargion, D., Grossi, M., De Sanctis Lucentini, P.~G., Di Troia, C.~
387: %``Clustering, anisotropy, spectra of ultra high energy cosmic ray:  
388: %Finger-prints of relic neutrinos masses in dark halos,''
389: 2001, Suppl. B, J.\ Phys.\ Soc.\ Jpn., 70, 46 
390: %%CITATION = HEP-PH 0108050;%%
391: 
392: \bibitem[Fargion (2002)]{Fargion2}
393: Fargion, D. 2002,
394: %``Discovering ultra high energy neutrinos by horizontal and upward 
395: %tau  air-showers: First evidences in terrestrial gamma flashes,''
396: \apj,  570, 909
397: %%CITATION = ASTRO-PH 0002453;%%
398: 
399: \bibitem[Fossati et\ al.(1998)]{acc-ab2}
400: Fossati, G., Maraschi, L., Celotti, A., Comastri, A., 
401: \& Ghisellini, G.\ 1998,
402: \mnras, 299, 433
403: 
404: \bibitem[Gorbunov et\ al.(2001)]{Gorbunov}
405: Gorbunov, D.~S., Raffelt G.~G. \& Semikoz, D.~V. 2001,
406: %``Axion-like particles as ultrahigh-energy cosmic rays?,''
407: \prd, {64}, 096005
408: %%CITATION = HEP-PH 0103175;%%
409: 
410: \bibitem[Grenier(2000)]{Grenier}
411: Grenier, I.~A. 2000, 
412: \aap, 364, L93
413: 
414: \bibitem[Ghi\-sel\-li\-ni et\ al.(1998)]{blazar-sequence} 
415: Ghisellini, E., Celotti, A., Fossati, G., Maraschi, L., \& 
416: Comastri, A. 1998,
417: \mnras, 301, 451
418: %%CITATION = ASTRO-PH 9807317;%%
419: 
420: \bibitem[Hartman et\ al.(1999)]{3EG} 
421: Hartman, R.~C. et\ al. 1999,
422: \apjs, 123, 79
423: 
424: \bibitem[Kalashev et\ al.(2001)]{Kalashev:2001qp}
425: Kalashev, O.~E., Kuzmin, V.~A., Semikoz, D.~V., \& Tkachev, I.~I. 2001,
426: astro-ph/0107130
427: 
428: \bibitem[Mattox et\ al.(2001)]{identifications} 
429: Mattox, J.~R., Hartman, R.~C., \& Reimer, O. 2001,
430: \apjs, 135, 155
431: 
432: \bibitem[Neronov et\ al.(2002)]{Neronov:2002se}
433: Neronov, A., Semikoz, D., Aharonian F., and Kalashev, O. 2002,
434: \prl, 89, 051101
435: %%CITATION = ASTRO-PH 0201410;%%
436: 
437: \bibitem[Punsly(1999)]{punsly} 
438: Punsly, B. 1999, 
439: \apj, 516, 141
440: 
441: \bibitem[Sigl et\ al.(2001)]{Sigl}
442: Sigl, G., Torres, D.~F., Anchordoqui, L.~A., \& Romero, G.~E. 2001,
443: \prd, 63, 081302
444: %%CITATION = ASTRO-PH 0008363;%%
445: 
446: \bibitem[Takeda et\ al.(1999)]{AG} 
447: Takeda, M.\ et al. 1999,
448: \apj, 522, 225
449: %[astro-ph/0008102].
450: 
451: \bibitem[Ti\-nyakov \& Tkachev(2001a)]{Tinyakov:2001ic}
452: Tinyakov, P.~G., \& Tkachev, I.~I. 2001a,
453: JETP Lett., 74, 1 (translation from 
454: Pisma Zh.\ Eksp.\ Teor.\ Fiz., 74, 3)
455: %%CITATION = ASTRO-PH 0102101;%%
456: 
457: \bibitem[Ti\-nyakov \& Tkachev(2001b)]{Tinyakov:2001nr}
458: Tinyakov, P.~G., \& Tkachev, I.~I. 2001b,
459: JETP Lett., 74, 445 (translation from
460: Pisma Zh.\ Eksp.\ Teor.\ Fiz., 74, 499)
461: %%CITATION = ASTRO-PH 0102476;%%
462: 
463: \bibitem[Ti\-nyakov \& Tkachev(2001c)]{Tinyakov:2001ir}
464: Tinyakov, P.~G., \& Tkachev, I.~I. 2001c,
465: astro-ph/0111305.
466: %%CITATION = ASTRO-PH 0111305;%%
467: 
468: \bibitem[Uchihori et\ al.(2000)]{clusters1} 
469: Uchihori, Y., Nagano, M., Takeda, M., Teshima, M., Lloyd-Evans, J.,
470: \& Watson, A.~A. 2000,
471: Astropart.\ Phys.,  13, 151
472: %[astro-ph/9908193];\\
473: %%CITATION = ASTRO-PH 9908193;%%
474: 
475: \bibitem[V\'eron-Cetty \& V\'eron(2001)]{Veron} 
476: V\'eron-Cetty, M.-P., \& V\'eron, P. 2001,
477: \aap, 374, 92
478: 
479: \bibitem[Weiler (1999)]{Weiler} Weiler, T. 1999,
480: Astropart.\ Phys.\  11, 30
481: %%CITATION = HEP-PH 9710431;%%
482: 
483: \end{thebibliography}
484: 
485: \newpage 
486: 
487: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
488: \begin{table}
489: \begin{tabular}{l|c|c|c|c|c|c|l}
490: \footnotesize 3EG J &\footnotesize E ID &\footnotesize Possible BLL      &\footnotesize l &\footnotesize b &\footnotesize z  &\footnotesize E &\footnotesize Q\\
491: \footnotesize (1) &\footnotesize (2) &\footnotesize (3) &\footnotesize (4) &\footnotesize (5) &\footnotesize (6) &\footnotesize (7) &  \footnotesize (8) \\
492: \hline
493: \footnotesize 0433+2908 &\footnotesize AGN &\footnotesize 2EG J0432+2910*   &\footnotesize  170.5 &\footnotesize  -12.6 &\footnotesize   ---  &\footnotesize  5.47 &\footnotesize 0,$\pm1$\\
494: \footnotesize  &\footnotesize &\footnotesize &\footnotesize  &\footnotesize   &\footnotesize     &\footnotesize   4.89  &\footnotesize 0,$+1$\\
495: \footnotesize 0808+5114 &\footnotesize AGN? &\footnotesize 1ES 0806+524*   &\footnotesize  166.2 &\footnotesize  32.91 &\footnotesize 0.138 &\footnotesize 3.4  &\footnotesize 0\\
496: \footnotesize  &\footnotesize  &\footnotesize    &\footnotesize      &\footnotesize  &\footnotesize  &\footnotesize 2.8  &\footnotesize 0\\
497: \footnotesize  &\footnotesize  &\footnotesize    &\footnotesize    &\footnotesize  &\footnotesize  &\footnotesize 2.5  &\footnotesize 0\\
498: \footnotesize 0812-0646 &\footnotesize AGN? &\footnotesize 1WGA J0816.0-0736 &\footnotesize  229.8 &\footnotesize  14.96 &\footnotesize  0.04 &\footnotesize --- &\footnotesize \\
499: \footnotesize 1009+4855 &\footnotesize AGN? &\footnotesize GB 1011+496     &\footnotesize  165.5 &\footnotesize  52.71 & \footnotesize   0.2 &\footnotesize --- &\footnotesize \\
500: \footnotesize 1052+5718 &\footnotesize AGN? &\footnotesize RGB J1058+564* &\footnotesize  149.6 &\footnotesize  54.42 &\footnotesize 0.144 &\footnotesize  7.76  &\footnotesize 0,$-1$\\
501: \footnotesize  &\footnotesize &\footnotesize &\footnotesize   &\footnotesize     &\footnotesize &\footnotesize  5.35  &\footnotesize 0,$-1$\\
502: \footnotesize  &\footnotesize &\footnotesize &\footnotesize   &\footnotesize     &\footnotesize &\footnotesize  5.50  &\footnotesize $-1$\\
503: \footnotesize 1222+2841 &\footnotesize AGN &\footnotesize ON 231*         &\footnotesize  201.7 &\footnotesize  83.29 &\footnotesize 0.102 &\footnotesize --- &\footnotesize \\
504: \footnotesize 1310-0517 &\footnotesize &\footnotesize 1WGA J1311.3-0521 &\footnotesize  312.1 &\footnotesize  57.16 &\footnotesize  0.16 &\footnotesize --- &\footnotesize \\
505: \footnotesize 1424+3734 &\footnotesize &\footnotesize TEX 1428+370   &\footnotesize  63.95 &\footnotesize  66.92 &\footnotesize 0.564 &\footnotesize  4.97  &\footnotesize 0,$+1$\\
506: \footnotesize 1605+1553 &\footnotesize AGN &\footnotesize PKS 1604+159*   &\footnotesize  29.38 &\footnotesize  43.41 &\footnotesize ---    &\footnotesize --- &\footnotesize \\
507: \footnotesize 1621+8203 &\footnotesize &\footnotesize 1ES 1544+820    &\footnotesize  116.5 &\footnotesize  32.97 &\footnotesize ---    &\footnotesize 2.7 &\footnotesize +1\\
508: \footnotesize 1733+6017 &\footnotesize &\footnotesize RGB J1742+597    &\footnotesize  88.46 &\footnotesize  31.78 &\footnotesize ---    &\footnotesize 2.5 &\footnotesize +1\\
509: \footnotesize  &\footnotesize &\footnotesize &\footnotesize      &\footnotesize  &\footnotesize     &\footnotesize 6.93 &\footnotesize $-1$\\
510: \footnotesize 1850+5903 &\footnotesize &\footnotesize RGB J1841+591    &\footnotesize  88.68 &\footnotesize  24.29 &\footnotesize  0.53 &\footnotesize 5.8 &\footnotesize +1\\
511: \footnotesize  &\footnotesize &\footnotesize  &\footnotesize     &\footnotesize  &\footnotesize  &\footnotesize 2.8 &\footnotesize +1\\
512: \footnotesize 1959+6342 &\footnotesize &\footnotesize 1ES 1959+650     &\footnotesize  98.0  &\footnotesize  17.67 &\footnotesize 0.047 &\footnotesize  5.5 &\footnotesize +1\\
513: \footnotesize 2352+3752 &\footnotesize AGN? &\footnotesize TEX 2348+360   &\footnotesize  109.5 &\footnotesize -24.91 &\footnotesize 0.317 &\footnotesize --- &\footnotesize \\
514: \end{tabular}
515: \caption{List of BL Lac objects associated with EGRET sources and
516: UHECR which contribute to correlations.}  {Note: (1) EGRET name; (2)
517: EGRET identification; (3) suggested BL Lac counterpart; the five
518: objects marked with as asterisk are the cases when suggested
519: identification argrees with the SIMBAD database; (4) and (5) Galactic
520: coordinates of the BL Lac counterpart; (6) redshift of the BL Lac
521: counterpart as given by \citet{Veron}; (7) energies of correlating
522: cosmic rays in units of $10^{19}$~eV; (8) UHECR charge assignments
523: under which the correlation occurs. }
524: \end{table}
525: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
526: 
527: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
528: \begin{table}
529: \begin{tabular}{l|c|c|c|c|c|c}
530: %\footnotesize 3EG J &\footnotesize E ID &\footnotesize Possible BLL
531: \footnotesize Q & 
532: \multicolumn{3}{c|}{\footnotesize antisymmetric field} & 
533: \multicolumn{3}{c}{\footnotesize symmetric field}  \\ 
534: \hline                 
535:        &\footnotesize  $p(\delta )$ &\footnotesize  $N_d(\delta)$  &\footnotesize  $\delta$ 
536:        &\footnotesize  $p(\delta )$ &\footnotesize  $N_d(\delta)$  &\footnotesize  $\delta$ \\
537: \hline
538: \footnotesize $0$    &\footnotesize  $ 10^{-4}$ &\footnotesize  8  &\footnotesize  $2.9^{\circ}$ 
539:        &\footnotesize  $ 10^{-4}$ &\footnotesize  8  &\footnotesize  $2.9^{\circ}$  \\
540: \footnotesize $+$    &\footnotesize  $7\cdot 10^{-5}$   &\footnotesize  8  &\footnotesize  $2.7^{\circ}$ 
541:        &\footnotesize  $9\cdot 10^{-4}$   &\footnotesize  9  &\footnotesize  $3.7^{\circ}$  \\
542: \footnotesize $0,+$  &\footnotesize  $3\cdot 10^{-7}$   &\footnotesize  13 &\footnotesize  $2.7^{\circ}$ 
543:        &\footnotesize  $2\cdot 10^{-6}$ &\footnotesize  12 &\footnotesize  $2.6^{\circ}$  \\
544: \footnotesize $0,\pm$&\footnotesize  $10^{-6}$ &\footnotesize  15 &\footnotesize  $2.8^{\circ}$ 
545:        &\footnotesize  $2\cdot 10^{-6}$ &\footnotesize  15 &\footnotesize  $2.9^{\circ}$
546: \end{tabular}
547: \caption{Summary of correlations between 14 BL Lac objects and 65
548: cosmic rays for different charge assignments and models of the GMF.}
549: \end{table}
550: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
551: 
552: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
553: \begin{table}
554: \begin{tabular}{l|c|c|c|l}
555:  \footnotesize 3EG J     &\footnotesize      l &\footnotesize      b &  \footnotesize E &  \footnotesize Q\\
556: % \footnotesize (1)       &\footnotesize (2) &\footnotesize      (3) &\footnotesize      (4) &  \footnotesize (5) \\
557: \hline
558:  \footnotesize 0245+1758 &\footnotesize 157.6  &\footnotesize -37.11 &\footnotesize 3.2   &\footnotesize +1\\ 
559:  \footnotesize 0329+2149 &\footnotesize 165.0  &\footnotesize -27.88 &\footnotesize 4.8   &\footnotesize +1\\ 
560:  \footnotesize 0429+0337 &\footnotesize 191.4  &\footnotesize -29.08 &\footnotesize 6.19   &\footnotesize 0,$+1$\\ 
561:  \footnotesize 1227+4302 &\footnotesize 138.6  &\footnotesize  73.33 &\footnotesize 4.3   &\footnotesize +1\\ 
562:  \footnotesize 1308+8744 &\footnotesize 122.7  &\footnotesize  29.38 &\footnotesize 3   &\footnotesize +1 \\
563:  \footnotesize 1337+5029 &\footnotesize 105.4  &\footnotesize  65.04 &\footnotesize 5.68   &\footnotesize +1\\ 
564:  \footnotesize 1621+8203 &\footnotesize 115.53 &\footnotesize  31.77 &\footnotesize 2.7&\footnotesize +1\\ 
565:  \footnotesize 1824+3441 &\footnotesize  62.49 &\footnotesize  20.14 &\footnotesize 9.79   &\footnotesize 0,$\pm1$ \\
566:  \footnotesize 1835+5918 &\footnotesize  88.74 &\footnotesize  25.07 &  5.8&\footnotesize +1\\ 
567:  \footnotesize  &\footnotesize   &\footnotesize   &  2.8&\footnotesize +1
568: \end{tabular}
569: \caption{List of unidentified EGRET sources correlating with cosmic
570: rays.}  {Note: Columns are the same as in Table 1.}
571: \end{table}
572: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
573: 
574: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
575: \begin{figure}
576: \plotone{f1.eps}
577: \caption{Significance of correlations between 14 $\gamma$-ray--loud BL
578: Lac objects and UHECRs as a function of the angular scale $\delta$ for
579: the $Q=0,1$ charge composition. This corresponds to the lowest
580: probability entry of Table 2.}
581: \label{SS-EV}
582: \end{figure}
583: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
584: 
585: \end{document}
586: