astro-ph0610944/art.tex
1: %&tex
2: \input magic
3: \input epsf
4: \let\label=\autoeqno
5: \let\ref=\eqref
6: 
7: \def\fig#1{\epsfxsize=0.49\hsize \epsfbox{#1.eps}}
8: \def\bfig#1{\epsfysize=0.355\vsize \epsfbox{#1.eps}}
9: \def\textbf#1{{\bf#1}}
10: \def\texttt#1{{\tt#1}}
11: \def\mbox#1{\hbox{#1}}
12: 
13: \hyphenation{an-a-lyse be-hav-iour}
14: \def\Ne{N_{\rm e}}
15: \def\Nmin{N_{\rm min}}
16: %--------------------------------------------------------------------------
17: \input refs
18: \preprint{astro-ph/0610944}
19: 
20: \title{Search for Sources of Cosmic Rays in the Region of the Knee.~II}
21: 
22: \author{G. V. Kulikov, M. Yu.\ Zotov}
23: \address{
24:    D. V. Skobeltsyn Institute of Nuclear Physics\\
25:    Moscow State University, Moscow 119992, Russia\\
26:    \tt $\{$kulikov,zotov$\}$@eas.sinp.msu.ru}
27: 
28: \date{October 31, 2006}
29: 
30: \bigskip
31: \centerline{\bf Abstract}
32: {\narrower\noindent
33: New results of an analysis of arrival directions of extensive air showers
34: registered with the EAS--1000 Prototype array from August 1997 till
35: February 1999 are presented. The method of Alexandreas et al., which has
36: been used for analysis of data registered with CYGNUS, Milagrito, HEGRA
37: AIROBICC, KASCADE and a number of other experiments, is employed. The
38: existence of zones of excessive flux of cosmic rays with energies in the
39: region of the knee is confirmed, as well as closeness of the zones to
40: coordinates of possible astrophysical cosmic ray sources.
41: 
42: }
43: 
44: %--------------------------------------------------------------------------
45: \section{Introduction}{sec:intro}
46:    We have already studied arrival directions of extensive air showers
47:    (EAS) with energies of primary particles of the order of
48:    $10^{14}$--$10^{15}$~eV registered with the EAS--1000 Prototype
49:    array~[\cref{we1},~\cref{we2}].
50:    In these works, the analysis was performed with the help of a special
51:    procedure of filtering experimental data which allowed us to obtain a
52:    set of data such that at some scale, the showers were distributed
53:    uniformly wrt.\ sidereal time and the azimuth angle.
54:    In this method, filtering of data was done in a (pseudo-)random way.
55:    Thus, in what follows, we shall call this procedure as a method of
56:    random filtering~(MRF).
57:    A detailed description of the MRF can be found in~[\cref{we1},~\cref{we2}].
58: 
59:    The MRF applied to our data set resulted in the selection of 37~zones
60:    of excessive flux (ZEFs) of EAS.
61:    Location of the majority of the ZEFs was found to be close to
62:    the coordinates of possible astrophysical sources of cosmic rays (CRs)
63:    with energies in the region of the knee in the energy spectrum
64:    at around 3~PeV~\cite{knee}.
65:    The fact that application of the MRF to the available data set
66:    led to exclusion of about~17\% of the initial number of EAS with
67:    known arrival directions, motivated us to perform a similar
68:    analysis basing on the method of Alexandreas et al.~\cite{Al}.
69:    This method utilizes the full amount of available data.
70:    It has been successfully employed earlier in a number of investigations,
71:    see, e.g., reports on the analysis of data obtained with the
72:    CYGNUS~\cite{Al}, Milagrito~\cite{Milagrito}, Milagro~\cite{Milagro},
73:    HEGRA AIROBICC~\cite{HEGRA}, and KASCADE~\cite{KASCADE} experiments.
74:    Below we present the results of our investigation.
75: 
76: %--------------------------------------------------------------------------
77: \section{Experimental Data}{sec:data}
78:    The data set under consideration includes 1,668,489 EAS registered
79:    during 203 days of operation of the EAS--1000 Prototype Array in the
80:    period from August~30, 1997, till February~1, 1999.
81:    The array consisted of eight detectors 
82:    located in the central part of the EAS MSU array along longer sides 
83:    of the $64\,{\rm m}\times22\,{\rm m}$ rectangle~\cite{Fomin}.
84:    The geographical coordinates of the array were
85:    $37^\circ32.5'$E, $55^\circ41.9'$N.
86: 
87:    Arrival directions were determined for 1,366,010 EAS.
88:    A number~$\Ne$ of charged particles (electrons) was found for 826,921
89:    EAS.
90:    The mean value $\bar\Ne=1.20\times10^5$ corresponds to the energy
91:    of a primary proton $E\approx10^{15}$~eV.
92:    There are only 11262 EAS with $\Ne>10^6$ but they give a noticeable
93:    contribution in the value of~$\bar\Ne$.
94:    With the above energetic EAS excluded, $\bar\Ne=7.65\times10^4$
95:    and the median value equals $3.62\times10^4$.
96:    Thus, the majority of showers in the data set under consideration
97:    had primary particles with energies slightly below the knee.
98: 
99: 
100: %--------------------------------------------------------------------------
101: \section{Method of Investigation}{sec:method}
102:    The main idea of the method of Alexandreas et al.~\cite{Al}
103:    is the following.
104:    A search for ZEFs is based on the comparison of real experimental data
105:    set with a ``background'' set that would be registered with the same
106:    installation and during the same period of observation in the
107:    assumption of a uniform distribution of arrival directions of EAS over
108:    the celestial sphere.
109:    The background data set is obtained by a simple procedure of
110:    randomization.
111:    First, for every shower with an arrival direction $(\theta,\varphi)$
112:    an arrival time~$t'$ of another
113:    shower is chosen randomly from the rest of the experimental data set.
114:    Next, arrival directions of all showers in the equatorial coordinates
115:    $(\alpha,\delta)$ are calculated basing on the respective triplets
116:    $(t',\theta,\varphi)$.
117:    In the resulting set, the distribution of EAS wrt.~$\delta$ is exactly
118:    the same as in the original data set.
119: 
120:    Randomization is performed multiple times in order to avoid
121:    a dependency of the distribution wrt.~$\alpha$ on the concrete
122:    choice of~$t'$.
123:    After this, one creates an averaged ``background'' map by
124:    calculating the mean number of EAS located in cells of a
125:    fixed size $\Delta\alpha\times\Delta\delta$ in all obtained maps.
126:    It is believed that the background map has most of the
127:    properties of an isotropic background~\cite{KASCADE}.
128:    A search for ZEFs is performed by comparing the number of EAS
129:    in a region in the original (``real'') experimental map
130:    with the same region in the background map.
131:    The measure of deviation is the significance
132:    $S=(N_{\rm real} - N_{\rm bg})/\sqrt{N_{\rm bg}}$, where
133:    $N_{\rm real}$ and $N_{\rm bg}$ are the numbers of EAS in
134:    the same region of the ``real'' experimental map and the
135:    background maps respectively.
136: 
137:    For the purposes of this investigation, the background map
138:    was obtained by averaging the number of showers binned in
139:    $\Delta\alpha\times\Delta\delta=1^\circ\times1^\circ$ cells.
140:    An analysis of the background maps revealed that two
141:    maps obtained consecutively differ by at most~1\% of the number
142:    of EAS in the respective cells after around~50 cycles of
143:    randomization, and by~0.5\% after~90 cycles.
144:    Thus we chose to perform 100 cycles of randomization
145:    to obtain the final background map.
146: 
147:    To look for ZEFs, we followed the same procedure as we used
148:    in~[\cref{we1}, \cref{we2}].  Namely, the field of observation was
149:    divided into strips of width $\Delta\delta=3^\circ\dots30^\circ$,
150:    with the boundaries of the strips of equal width shifted by~$1^\circ$
151:    wrt.\ each other.
152:    Each strip was then divided into cells of equal width~$\Delta\alpha$.
153:    In the main part of the investigation, for any fixed width~$\Delta\delta$
154:    of a strip, $\Delta\alpha$ was chosen to be equal to a rounded value
155:    (expressed in degrees) of $\Delta\delta/\cos\bar{\delta}$, where~$\bar{\delta}$
156:    is the mean value of the declination for the strip.
157:    This guarantees that cells with the same~$\Delta\delta$ but located
158:    at different declination, have an approximately equal area.
159:    Similar to~[\cref{we1}, \cref{we2}], such cells will be called `regular'.
160:    It is important to take into account that neither the method
161:    of Alexandreas et al., nor the method of random filtering put
162:    a restriction on the way one chooses regions to be compared.
163:    One may perform a search for ZEFs with an arbitrary relation
164:    of~$\Delta\alpha$ and~$\Delta\delta$, as well as ZEFs of an
165:    arbitrary shape.
166: 
167: 
168: %--------------------------------------------------------------------------
169: \section{Main Results}{sec:main}
170:    Application of the method of Alexandreas et al. to the available data
171:    set resulted in the discovery of~561 regular cells with $S>3.0$.
172:    For 364 of these cells, significance $S>3.1$.
173:    For the sake of convenience, the cells were joined into 27 zones
174:    such that each zone is a connected domain.
175:    In order to perform an accurate comparison of the results of the two
176:    methods, we added five zones made of ``irregular'' cells, i.e., cells
177:    with an arbitrary relation of~$\Delta\alpha$ and~$\Delta\delta$.
178:    The resulting ZEFs are shown in Fig.~1 in red.
179:    The most of partially overlapping cells of excessive flux
180:    are not shown for the sake of visual clarity.
181:    Their joint exterior boundaries are shown instead.
182:    Blue lines show the boundaries of the ZEFs presented
183:    in~[\cref{we1}, \cref{we2}].  Recall that all of these 37 ZEFs
184:    but zones No.~4, 5, 6, 20, and~21, and bigger cells in ZEFs
185:    No.~3, 27, 29, 33, and~36 are composed of regular cells.
186:    A number of the zones found by the MRF coincide with those
187:    found by the method of Alexandreas et al.\ and thus their
188:    boundaries are not visible in the figure.
189: 
190: %--------------------------------------------------------------------------
191:    \figure{fig:zefs}{Red lines show the boundaries of ZEFs found by the
192:    method of Alexandreas et al. Blue lines show the boundaries of ZEFs
193:    found by the method of random filtering. Numbers in the field of the
194:    figure show the numbers of ZEFs assigned to them in~[\cref{we1},
195:    \cref{we2}]. The (magenta) arcs show the Galactic plane. The $\cap$-like
196:    curve shows the Supergalactic plane.}{\bfig{zefs}}
197: %--------------------------------------------------------------------------
198: 
199:    As is clear from Fig.~1, the results obtained by two different methods
200:    are quite close, though the MRF seems to be more selective in the
201:    sense that the total area of ZEFs found by this method is less than
202:    that of the zones found by the method of Alexandreas et al.\
203:    (``A-zones''). In sum, 20 ZEFs found by the MRF lie {\it inside\/}
204:    A-zones. These are zones No.~1--9, 12, 14, 15, 18, 22, 25--28, 31,
205:    and~36. Zone No.~30 also lies inside the corresponding A-zone
206:    excluding a small area
207:    $\Delta\alpha\times\Delta\delta=3^\circ\times1^\circ$. A major part
208:    of ZEF No.~32 also lies within the corresponding A-zone. Zones No.~11,
209:    17, and~21 exactly coincide with A-zones, while zones No.~13 and
210:    33--35 coincide with A-zones up to~$1^\circ$. In fact, there are
211:    multiple coincidences with deviation of at most~$1^\circ$ of separate
212:    cells found my the two methods  in all cases when a ``composite'' ZEF
213:    found by the MRF is embedded into a zone found by the method of
214:    Alexandreas et~al.
215: 
216:    Let us mention the existence of pairs for the central cell in ZEF
217:    No.~16, which has the M33 galaxy located inside, and for ZEF No.~7,
218:    which contains the famous C2 triplet of ultra-high energy cosmic rays,
219:    registered with the AGASA array~\cite{AGASA}. Recall that the
220:    observation of the latter coincidence was one of the motivations for
221:    the analysis of coordinates of extra-galactic cosmic ray sources
222:    performed in~\cite{we2}.
223: 
224:    An embedding of A-zones into the ZEFs found by the MRF can be seen in
225:    a few cases. Namely, ZEF No.~10, 23, 24, and~29 contain the
226:    corresponding A-zones inside. Only three of 37 ZEFs found by the MRF
227:    do not have close counterparts among A-zones. These are the irregular
228:    ZEF No.~20, the irregular (exterior) cell of ZEF No.~29, and ZEF
229:    No.~37.  In the latter case, there are a number of irregular A-zones
230:    that intersect the ZEF.  They are shown in the figure.
231: 
232:    Finally, notice seven A-zones (made of regular cells) that do not
233:    have any close counterparts among the ZEFs found by the MRF.
234:    These are a zone located around $\alpha=20^\circ,\,\delta=75^\circ$,
235:    two big A-zones in the vicinity of $\delta=70^\circ$,
236:    a zone near $\alpha=330^\circ,\,\delta=73^\circ$,
237:    a zone around $\alpha=290^\circ,\,\delta=60^\circ$,
238:    and two zones near $\delta=10^\circ$ ($\alpha=25\dots29^\circ$ and
239:    $\alpha=224\dots229^\circ$ respectively).
240: 
241: %--------------------------------------------------------------------------
242: \section{Discussion}{sec:disc}
243:    In our opinion, there is a qualitative agreement between the results
244:    of the analyses of arrival directions of cosmic rays registered with
245:    the EAS--1000 Prototype Array performed by two different methods.
246:    The agreement confirms one of the main conclusions
247:    of~[\cref{we1}, \cref{we2}] that there are zones of excessive flux
248:    of EAS in the available data set.
249:    We find it remarkable that mutual coincidences or embeddings are
250:    observed for all zones located in the vicinity of the Galactic
251:    and Supergalactic planes.
252: 
253: %--------------------------------------------------------------------------
254: \figure{fig:objects}{Zones of excessive flux found by the
255:    method of Alexandreas et~al.~\cite{Al} that have counterparts
256:    among the ZEFs found in~[\cref{we1}, \cref{we2}].
257:    Different symbols show coordinates of some astrophysical objects
258:    that have coordinates within angular distance $\le3^\circ$
259:    from the ZEFs.  These are galactic supernova remnants~$(\bullet)$,
260:    pulsars (filled blue triangles), active galactic nuclei and 
261:    interacting galaxies at red shifts $z\le0.01$~($\triangle$).
262:    Also shown are the coordinates of the C2 triplet registered
263:    with the AGASA array~($*$)~\cite{AGASA} and the M33
264:    galaxy (the red box).}{\bfig{objects}}
265: %--------------------------------------------------------------------------
266: 
267:    Let us now consider correlations between positions of the ZEFs
268:    and coordinates of possible astrophysical sources of cosmic rays
269:    with energies near the knee.
270:    Figure~2 shows A-zones that have close counterparts among the ZEFs
271:    presented in~[\cref{we1}, \cref{we2}] and coordinates of galactic
272:    supernova remnants (SNRs)~\cite{Green}, pulsars~\cite{ATNF}
273:    (as of the state on June~19, 2006), and active galactic nuclei (AGN)
274:    and interacting galaxies at red shifts $z\le0.01$~\cite{Simbad}.
275:    All shown astrophysical objects have coordinates at angular
276:    distances~$\le3^\circ$ from the closest A-zone.
277:    This criterion selected 18 of 110 SNRs at $\delta>-16^\circ$
278:    with 15 of them lying inside A-zones and two of them lying
279:    at angular distances~$\le1.5^\circ$,
280:    154 of 694 pulsars, and 139 of 317 AGN and interacting galaxies.
281:    In particular, the selected objects include the famous SNRs
282:    SN1181, 3C461 (Cas~A), SN1572 (Tycho), and the Crab Nebula,
283:    a number of pulsars located at distances~$<0.5$~kpc from the Solar
284:    system, as well as the active galactic nuclei 
285:    NGC 3718, NGC 2681, NGC 4278, Mrk 1307, M105, M64, M65, M66, etc.,
286:    which have red shifts $z\le0.0033$.
287:    Notice that the Virgo cluster of galaxies lies inside a huge A-zone 
288:    located along the Supergalactic plane 
289:    ($\alpha\approx170^\circ\dots200^\circ$).
290:    It is evident that there are numerous coincidences between
291:    coordinates of possible astrophysical sources of cosmic rays
292:    with energies near the knee and positions of zones located
293:    in the vicinity of the Galactic and Supergalactic planes.
294: 
295: %--------------------------------------------------------------------------
296: \asection{Acknowledgments}
297:    Only free, open source software was used for the investigation.
298:    In particular, all calculations were performed with
299:    GNU Octave~\cite{Octave} running in Slackware Linux.
300:    The research was partially supported by the Federal Program
301:    ``Priority Investigations in Science and Technology'' for
302:    2002--2006, contract No.~02.452.11.7053, and the Russian
303:    Foundation for Fundamental Research grant No.~05-02-16401.
304: 
305: 
306:    
307: \listrefs
308: \bye
309: %--------------------------------------------------------------------------
310: 
311: 
312: