1: %%
2: %
3: %
4: % A Harmonic Analysis of the Large Scale Cosmic Ray Anisotropy
5: % Author: Brian Kolterman
6: %
7:
8: %Class Required
9: \documentclass[12pt,preprint]{aastex}
10:
11: %The paper title
12: \title{The Large Scale Cosmic-Ray Anisotropy as Observed with Milagro}
13:
14: %Short title to print in the headers to the final publication (Not showed in this print).
15: \shorttitle{Cosmic Ray Anisotropy}
16: %All paper authors
17: %\author{B. E. Kolterman$^{1}$ for the Milagro Collaboration.}
18:
19: \author{A. A. Abdo\altaffilmark{1,2}, B. T. Allen\altaffilmark{3},
20: T. Aune\altaffilmark{4}, D. Berley\altaffilmark{5}, S. Casanova\altaffilmark{6},
21: C. Chen\altaffilmark{3}, B. L. Dingus\altaffilmark{6}, \\ R. W. Ellsworth\altaffilmark{7}, L. Fleysher\altaffilmark{8},
22: R. Fleysher\altaffilmark{8}, M. M. Gonzalez\altaffilmark{9},
23: J. A. Goodman\altaffilmark{5},\\ C. M. Hoffman\altaffilmark{6}, B. Hopper\altaffilmark{5}, P. H. H\"{u}ntemeyer\altaffilmark{6}, B. E. Kolterman\altaffilmark{8}, C. P. Lansdell\altaffilmark{5},\\ J. T. Linnemann\altaffilmark{10}, J. E. McEnery\altaffilmark{11}, A. I. Mincer\altaffilmark{8}, P. Nemethy\altaffilmark{8}, D. Noyes\altaffilmark{5}, J. Pretz\altaffilmark{6},\\ J. M. Ryan\altaffilmark{12}, P. M. Saz Parkinson\altaffilmark{4}, A. Shoup\altaffilmark{13}, G. Sinnis\altaffilmark{6}, A. J. Smith\altaffilmark{5}, G. W. Sullivan\altaffilmark{5},\\ V. Vasileiou\altaffilmark{5}, G. P. Walker\altaffilmark{6}, D. A. Williams\altaffilmark{4}, and G. B. Yodh\altaffilmark{3}}
24:
25: %Short title to print in the headers to the final puplication (Not showed in this print).
26: \shortauthors{A. A. Abdo and et al}
27: %All the affiliations.
28:
29:
30: \altaffiltext{1}{National Research Council Research Associate, National Academy of Sciences, Washington, DC 20001}
31: \altaffiltext{2}{Space Science Division, Naval Research Laboratory, Washington, DC 20375}
32: \altaffiltext{3}{Department of Physics and Astronomy, University of California, Irvine, CA 92697}
33: \altaffiltext{4}{Santa Cruz Institute for Particle Physics, University of California, 1156 High Street, Santa Cruz, CA 95064}
34: \altaffiltext{5}{Department of Physics, University of Maryland, College Park, MD 20742}
35: \altaffiltext{6}{Group P-23, Los Alamos National Laboratory, P.O. Box 1663, Los Alamos, NM 87545}
36: \altaffiltext{7}{ Department of Physics and Astronomy, George Mason University, 4400 University Drive, Fairfax, VA 22030}
37: \altaffiltext{8}{Department of Physics, New York University, 4 Washington Place, New York, NY 10003}
38: \altaffiltext{9}{ Instituto de Astronom\'{i}a, Universidad Nacional Aut\'{o}noma
39: de M\'{e}xico, D.F., M\'{e}xico, 04510}
40: \altaffiltext{10}{Department of Physics and Astronomy, Michigan State University,
41: 3245 BioMedical Physical Sciences Building, East Lansing, MI 48824}
42: \altaffiltext{11}{NASA Goddard Space Flight Center, Greenbelt, MD 20771}
43: \altaffiltext{12}{Department of Physics, University of New Hampshire, Morse Hall, Durham, NH 03824}
44: \altaffiltext{13}{The Ohio State University, Lima, OH 45804}
45:
46:
47: %\email{bek218@nyu.edu}
48:
49: %The abstract.
50: \begin{abstract}Results are presented of a harmonic analysis of the large scale
51: cosmic-ray anisotropy as observed by the Milagro observatory. We show a
52: two-dimensional display of the sidereal anisotropy projections in right
53: ascension generated by the fitting
54: of three harmonics to 18 separate declination bands. The Milagro
55: observatory is a water Cherenkov detector located in the Jemez mountains
56: near Los Alamos, New Mexico. With a high duty cycle
57: and large field-of-view, Milagro is an excellent instrument for measuring
58: this anisotropy with high sensitivity at TeV energies. The analysis is
59: conducted using a seven year data sample consisting of more than 95 billion
60: events, the largest such data set in existence. We observe an anisotropy with
61: a magnitude around 0.1\% for cosmic rays with a median energy of 6 TeV. The
62: dominant feature is a deficit region of depth (2.49 $\pm$ 0.02 stat.\ $\pm$
63: 0.09 sys.) $\times 10^{-3}$ in the direction of the Galactic North Pole
64: centered at 189 degrees right ascension. We observe a steady increase in the
65: magnitude of the signal over seven years.
66: \end{abstract}
67:
68: \keywords{Milagro --- Cosmic Rays --- Anisotropy --- Galactic Magnetic Fields --- Heliosphere}
69:
70: %\email{aastex-help@aas.org}
71:
72:
73: %%%%%%%%%%%%%%%%%%%% B E G I N D O C U M E N T%%%%%%%%%%%%%%%%%%%%%%%
74: \begin{document}
75: \maketitle
76: %Begin the section.
77:
78:
79: \section{Introduction}
80:
81: Observation of the sidereal large scale cosmic-ray (CR) anisotropy at
82: energies of 1 - 100 TeV is a useful tool in probing the magnetic field
83: structure in our interstellar neighborhood as well as the distribution of
84: sources. Cosmic-rays at these energies are almost entirely of Galactic origin
85: and are expected to be nearly isotropic
86: due to interactions with the Galactic magnetic field (GMF) \citep{AsCR}.
87: The gyro radii of CRs at these energies in a GMF of about $1\mu$G are from
88: about 100 AU - 0.1 pc which is much smaller than the size of the Galaxy. This
89: has the effect of trapping the CRs in the Galaxy for times on the order of
90: $10^{6}$ years. Inhomogeneities in the GMF are interspersed randomly
91: throughout the galaxy and in effect act as
92: scattering centers for CRs. This randomizes their directions as they propagate
93: leading to a high degree of isotropy on scales of a few hundred
94: parsecs.
95:
96: Anisotropy can be induced through both large scale and local magnetic
97: field configurations which cause deviations from the isotropic diffusion
98: approximation. At lower energies of around a TeV, the heliosphere
99: may be able to induce a CR excess in the direction of the heliotail
100: and also could modulate the overall CR anisotropy \citep{NgLC,NgTI}. At higher
101: energies, the contribution of discrete CR sources has been shown to be capable
102: of creating a large scale anisotropy \citep{PtSNR,SMP}.
103:
104: Diffusion of CRs out of the Galactic halo can also produce an anisotropy. Since
105: the matter density is higher in the Galactic disk compared to that in
106: the surrounding halo, the diffusion coefficient will generally be much higher
107: in the halo. For this reason, CRs produced in the Galactic disk will tend
108: to diffuse out into the halo creating an anisotropy in the direction
109: perpendicular to the disk. Predictions of the anisotropy have been made
110: using values of the diffusion coefficient inferred using a given propagation model and
111: observational data. This predicted anisotropy can take on values
112: of $10^{-2}$ to $10^{-5}$ depending on the propagation model used \citep{AsCR}. Given this correlation between the anisotropy and
113: diffusion coefficient, knowledge of the large scale CR anisotropy can be used
114: to constrain diffusion models.
115:
116: In addition to the above effects, Compton and Getting introduced a
117: theory \citep{CG} of CR anisotropy which predicts a dipole effect due to the
118: motion of an observer
119: with respect to an isotropic CR plasma rest frame. The anisotropy arising
120: from the Earth's motion around the sun is calculated to be on the order
121: of $10^{-4}$ and would appear in universal time. This effect has been
122: observed by numerous experiments (e.g. see \cite{ET}, \cite{TBCG}). There is also a possible sidereal time
123: anisotropy coming from the motion of the Solar System through the
124: Galaxy. This effect is more difficult to predict given that the isotropic
125: CR rest frame is not known. Using the assumption
126: that the CR rest frame does not co-rotate with the Galaxy leads to a predicted
127: anisotropy on the order of $\sim0.1\%$. This effect has not
128: been observed \citep{TB}.
129:
130: There have been numerous observations of the large-scale sidereal anisotropy
131: in the range of energies $10^{11} - 10^{14}$ eV. In this paper we use large-scale
132: anisotropy to refer to features of the anisotropy in the sky with an extent greater than
133: $\sim 40^{\circ}$ in right ascension. Most of these observations are examined by fitting
134: harmonics to the distribution of cosmic-ray events in the right ascension
135: direction. The main features of the anisotropy, the amplitude and phase, are
136: fairly constant over the energy range mentioned above ($3-10 \times 10^{-4}$
137: for the amplitude, and $0 - 4$ hr for the phase, see \cite{SK} for a
138: compilation). At TeV energies, the focus of this paper, the anisotropy
139: is known to have a value on the order of $10^{-3}$ with a deficit region,
140: sometimes called the ``loss-cone", around $200^{\circ}$ right
141: ascension, and an excess, or ``tail-in" region, around $75^{\circ}$ \citep{NgLC,NgTI}.
142: In this paper, we present the results of a harmonic analysis of the large scale
143: cosmic-ray anisotropy in the northern hemisphere as observed by the Milagro experiment.
144:
145: \section{The Milagro Observatory}
146:
147: The Milagro observatory \citep{Milagro} is a water Cherenkov detector which
148: is used to monitor extensive air showers produced by TeV gamma-rays and
149: hadrons hitting the Earth's atmosphere. Milagro is located in New Mexico at a
150: latitude of $35.88^{\circ}$, a longitude of $106.68^{\circ}$, and an altitude of 2630 m above sea level,
151: possessing a large field of view
152: of $\sim$2sr and a high duty factor of $>90$\%. The detector is composed of
153: an 80m$\times$60m$\times$8m pond filled with $\sim 23$ million liters of
154: purified water and protected by a light-tight cover. The central
155: pond is instrumented with two layers of PMTs: a top layer with 450 PMTs
156: under 1.4m of water which detects Cherenkov light from air shower
157: electrons, electrons Compton scattered by gamma-rays, and gamma-rays that have
158: converted to electron-positron pairs in the water; and a bottom layer with 273
159: PMTs 6m under the surface used for gamma-hadron separation (not used in
160: this analysis). The direction of an air shower is reconstructed
161: using the relative timing of the PMTs hit in the top layer of the pond with
162: an angular resolution of $<1^{\circ}$.
163:
164: This pond is surrounded by a 200m$\times$200m array of 175 ``outrigger"
165: tanks. Each ``outrigger" is a cylindrical, polyethylene tank with a
166: diameter of 2.4 m and a height
167: of 1 m. The outrigger tanks contain $\sim 4000$ liters of water and are
168: instrumented with a single downward facing PMT located at the top of the
169: tank. The inside of each tank is also lined with Tyvek (a white, reflective
170: material) to increase the light collection capability of the
171: PMT. The outrigger array, completed in 2003, is used to improve the angular
172: resolution and in the estimation of primary particle energy.
173:
174: The data used in this analysis were collected by Milagro from
175: July 2000 through July 2007. The hardware trigger was essentially a voter
176: coincidence, requiring a multiplicity of about 75 PMTs. During this time
177: the trigger multiplicity was tuned to maintain an average trigger rate of $\sim$1700 events per second
178: (to within 10 \%), the majority of which are due to hadronic showers. After event reconstruction,
179: accepted events are required to have used at least 50 PMTs in the angular fit
180: ($N_{fit}$) and have a zenith arrival angle of $\leq50^{\circ}$. The $N_{fit} \geq 50$ requirement
181: removes effects due to fluctuations in the trigger
182: threshold as events passing this cut generally have a higher PMT
183: multiplicity than the hardware trigger requires. Although this makes the data more stable, it also
184: reduces the effective trigger rate to $\sim 500$ Hz. After these cuts the
185: data set consists of $9.59\times10^{10}$ cosmic-ray events with a median
186: energy of 6 TeV as determined from the measured primary cosmic-ray spectra
187: and simulated detector efficiency. The median energy varies slowly with zenith
188: angle; it is 4 TeV for zenith angles from $5^{\circ}$ to
189: $10^{\circ}$, and 7 TeV for zenith angles from $35^{\circ}$ to
190: $40^{\circ}$.
191:
192:
193: Detector response to primaries is determined with a Monte Carlo simulation. This
194: Monte Carlo has two main components: the simulation of an air shower in the
195: atmosphere and the simulation of the shower through the Milagro detector. These
196: simulations are carried out using the CORSIKA \citep{CSKA} and
197: GEANT4 \citep{GEANT} packages respectively. Cosmic-ray nuclei ranging in
198: mass from the dominant protons to those as heavy as iron
199: are simulated according to their spectra as measured by the ATIC
200: experiment \citep{ATIC}. Because of the large shower to shower variations, Milagro
201: cannot measure the energy of individual primaries generating these air showers without
202: careful data selection criteria. Measurement
203: of energy correlated parameters does allow some determination of primary energy
204: spectra without severely restricting the data. The energy analysis of the large-scale
205: anisotropy will be the subject of a future publication.
206:
207:
208:
209:
210: \section{Data Analysis}
211:
212:
213: \subsection{Forward-Backward Asymmetry and Harmonic Method}
214: \label{fbahm}
215:
216: We have designed an analysis method to search for fractional deviations
217: from isotropy down to levels of about $10^{-4}$, below the level of
218: predictions of the Compton-Getting effect and the magnitude of previous
219: observations. Such sensitivity cannot be achieved by looking at
220: raw event rates as a function of direction, because the air shower development
221: essentially makes the atmosphere part of the detector, and weather effects
222: lead to typical overall trigger rate variations of $\pm 15 \%$. Thus, counting
223: raw event rates would give random apparent anisotropies
224: completely overshadowing a true signal. Instead, we employ the technique of
225: forward-backward asymmetry (FB) using the number of events ($N_{F}$
226: and $N_{B}$) collected in some small time interval in two ``telescopes", i.e.\
227: pixels of small and equal solid angle, at the same forward and backward
228: angle as shown in Figure \ref{fig1}.
229:
230: \begin{equation}
231: FB = (N_{F} - N_{B})/(N_{F} + N_{B})
232: \end{equation}
233:
234: The expression of FB is manifestly independent of overall
235: detector rate, as can be seen from the invariance of FB under the
236: substitution $N_{F}, N_{B} \rightarrow cN_{F}, cN_{B}$, where $c$ is
237: a constant.
238:
239:
240: The analysis method utilizes the rotation of the Earth to search for a
241: coherent modulation of FB during a 24 hour day. The FB modulation does
242: not directly give the anisotropy of the sky, but is a function of it. As
243: a region of the sky with an excess cosmic-ray flux relative to the average
244: (a positive anisotropy region) is swept into the forward ``telescope" FB
245: becomes more positive; when the same excess is swept into the backward
246: ``telescope", FB becomes negative. One can think of FB as a coarse
247: ``derivative" of the actual anisotropy of the sky. Thus the FB modulation
248: is a tool to obtain the quantity of interest, the fractional anisotropy
249: (i.e.\ deviation from uniformity) of the sky, as detailed below.
250:
251: The FB method is commonly used in particle physics (e.g.\ see \cite{FBCA})
252: and is also closely related to the East-West technique, classically
253: used in other anisotropy measurements (for an example see \cite{NgLC}), in
254: which the ratio examined for modulation is
255: $EW = (N_{E} - N_{W})/N_{tot}$, formed with the rate integrated over the
256: whole Eastern and Western sky. Again this is a rate independent quantity which
257: is used to remove random fluctuations in overall detector rate, a common
258: requirement in all experiments using the atmosphere as part of their
259: detector. The EW method can be thought of as a single integrated
260: measurement, compared to
261: our multiple localized and independent measurements of FB which are
262: used to improve the statistical power of this analysis.
263:
264: Random daily weather-induced or instrumental variations of the anisotropy
265: itself (and therefore of FB) are averaged out by summing (i.e.\
266: averaging) many full days, each sweeping a full circle in the sky. The
267: random anisotropies decrease as $\sqrt{N_{days}}$ while a coherent signal
268: is not reduced. Finally, because the method uses the time modulation of
269: FB rather than FB itself, it is impervious to the inherent geometrical
270: anisotropy present in the detector acceptance seen in the azimuthal
271: distribution of Figure \ref{fig2}, which was observed to be stable
272: during the lifetime of the experiment. We note that only instabilities
273: of this geometrical anisotropy on the scale of a single day would
274: affect the measurement, while a slow evolution would not.
275:
276: Because this method measures the modulation in the direction of the Earth's rotation, it
277: yields no information about the modulation in the declination (dec.)
278: direction. The results will therefore be for the projection of the
279: anisotropy in the right ascension (r.a.) direction rather than the full
280: 2-D anisotropy of the sky. Such a projection can be created for
281: any visible dec.\ band. Since each
282: dec.\ band sweeps around a full circle of the sky independently
283: and contains statistically independent data, we choose to do separate
284: analyses for each $5^{\circ}$ dec.\ band of our data, considering each
285: as a separate observation. Any theoretical description of the true 2-D
286: anisotropy can be confronted with, or constrained by, our data (given in
287: Section 4, Table 1) by projecting the anisotropy along r.a.\ in our
288: dec.\ bands. In the rest of this paper we will always use the word anisotropy
289: to mean this projected anisotropy.
290:
291: To do the harmonic fit we make the assumption that the large scale anisotropy
292: in any given dec.\ band can be modeled by a Fourier series and that it is a
293: small modulation of a nearly isotropic signal. Three harmonics
294: (the fundamental and the next two) have been found to be sufficient for this
295: method (see section 4.1). This allows us to see large scale effects having a
296: width in r.a.\ of greater than $\sim40^{\circ}$. In this harmonic model the
297: normalized rate in a given direction of the sky, in celestial equatorial
298: coordinates with declination = $\delta$ and right ascension = $\theta$ (or
299: the equivalent coordinate in the right ascension direction when using
300: a different coordinate system) is:
301:
302: \begin{equation}\label{skymod}
303: \frac{R_{\delta}(\theta)}{<R_{\delta}(\theta)>} = 1 + A_{\delta}(\theta) = 1+\sum_{n=1}^3 \gamma_{n,\delta} \cos n(\theta - \phi_{n,\delta}) \\
304: \end{equation}
305:
306: where for a given $\delta$ the average in the denominator of the left side is
307: over $\theta$, and $A_{\delta}(\theta)$ is the fractional anisotropy in the
308: right ascension direction. This model of the sky is complete once the
309: six Fourier coefficients (three amplitudes and three phases) on the
310: right side of the equation are known. The data analysis thus consists
311: of determining these Fourier coefficients.
312:
313: It is this harmonic model that allows us to connect the quantity of interest,
314: namely the fractional anisotropy in the sky $A_{\delta}(\theta)$, and our
315: chosen tool, the measurement of the modulation of the forward-backward
316: asymmetry. This connection is detailed explicitly below.
317:
318:
319:
320:
321: The quantity $N_{\theta_{0},\delta}(\xi)$ is the number of cosmic-ray events
322: collected during a particular time interval (characterized
323: by the angle $\theta_{0}$), in an angular bin at a given
324: declination ($\delta$) and local hour angle ($\xi$) characterizing
325: the symmetric forward ($+ \xi$) and backward ($- \xi$) inclination of the
326: ``telescope" of Fig.\ \ref{fig1}. In this notation the forward-backward asymmetry of Eq.
327: 1 becomes:
328:
329: \begin{equation}\label{eqn1}
330: FB_{\delta}(\theta_{0}, \xi) = \frac{N_{\theta_{0},\delta}(+\xi) - N_{\theta_{0},\delta}(-\xi)}{N_{\theta_{0},\delta}(+\xi) + N_{\theta_{0},\delta}(-\xi)}
331: \end{equation}
332:
333: The bin counts are computed for
334: $1/2$-hour intervals in sidereal time (ST), universal time (UT), and anti-sidereal time (AST) \citep{AST}. These $1/2$-hour intervals are parameterized by an angle
335: $\theta_{0}$ which specifies the relative advance of the local meridian
336: through the sky for the three different time scales:
337:
338: \[ \theta_{0} = 3.75^{\circ} + 7.5^{\circ}\times IST \]
339: \[ \theta_{0} = 3.75^{\circ} + 7.5^{\circ}\times IUT \]
340: \[ \theta_{0} = 3.75^{\circ} + 7.5^{\circ}\times IAST \]
341:
342: IST, IUT, and IAST are integers, from zero
343: to 47, denoting half hour intervals of sidereal time, UT, and anti-sidereal
344: time (defined by flipping the sign of the transformation from universal
345: time to sidereal time). In the above equations, the constants convert an
346: integration time interval (1/2 hour) into degrees ($7.5^{\circ}$) and a
347: starting angle at the center of that interval.
348:
349: In practice the cosmic-ray events are recorded in histograms with
350: $5^{\circ}\times5^{\circ}$ bins according to their arrival
351: direction from $-10^{\circ}$ to $80^{\circ}$ in declination and $-50^{\circ}$
352: to $+50^{\circ}$ in hour angle. The events are collected over a 30 ``minute"
353: period giving us 48 half hour histograms per day, where ``minute" is defined in the
354: three following time frames: sidereal
355: (366.25 days/year), universal (365.25 days/year) and anti-sidereal
356: (364.25 days/year). The coordinate system in the sidereal time frame is
357: ``sky-fixed", in the universal time frame it is
358: ``sun-fixed", and in the anti-sidereal frame it is ``non-physical". Histograms
359: for a given half hour period are accumulated over any number of days to build
360: an average set which corresponds to a chosen time period. A
361: representative example is shown in Figure \ref{fig3}. These 48 half hour
362: histograms for the time frame, period, and dec.\ band of interest are then
363: analyzed using the method of forward-backward asymmetry described above
364: combined with a harmonic fit.
365:
366:
367:
368:
369:
370: Recognizing that $\theta = \theta_{0} \pm \xi$, substituting Eq. \ref{skymod}
371: for the harmonic sky model into Eq. \ref{eqn1}, and then utilizing $\gamma \ll 1$
372: and the appropriate trigonometric identities, we get:
373:
374: \begin{equation}\label{fitfun}
375: FB_{\delta}(\theta_{0},\xi)\approx \sum_{n=1}^3 - \gamma_{n, \delta} \sin (n\xi) \sin (n(\theta_{0}-\phi_{n, \delta}))\nonumber\end{equation}
376:
377: Equation (\ref{fitfun}) is the prediction of the harmonic model of the sky,
378: in terms of its Fourier coefficients, for the forward-backward asymmetry, to
379: be fit with our experimental data.
380:
381: The next step is preparing the experimental data for the fit. For a fixed
382: ``telescope" angle $\xi$ and at a given $\delta$ we calculate FB and
383: its statistical error by using the relevant pair of histogram bins in
384: each of our 48 saved histograms of hour angle vs. dec.\ (see Fig. \ref{fig3}
385: for an example). Since we are only interested in the modulation of FB with
386: $\theta_{0}$ (i.e.\ with time), the average over all $\theta_{0}$
387: is subtracted from the calculated FB. This step
388: separates FB due to the inherent geometric anisotropy of the detector
389: from the time modulated FB due to the sky anisotropy. These values are
390: assembled into a 48 point, one-dimensional
391: histogram of FB and its error vs.\ $\theta_{0}$ for this particular
392: choice of $\xi$ and
393: $\delta$. Equation \ref{fitfun} could be fit to this histogram to
394: obtain the Fourier coefficients. But, this procedure can be done for
395: several choices of $\xi$ that we call $\xi_{i}$, each $\xi_{i}$ is
396: sampling the same sky, yielding a new FB vs.\ $\theta_{0}$ graph to fit. By
397: choosing a set of $\xi_{i}$ (where $\xi_{i}$ ranges from $2.5^{\circ}$ to a
398: dec.\ dependent maximum of up to $47.5^{\circ}$ in $5^{\circ}$ steps), centered on each of the
399: $5^{\circ}\times5^{\circ}$ pixels (see Fig. \ref{fig1} \& \ref{fig3}) and
400: calculating the FB (Eq.\ 3) with the contents of the corresponding pixels, we
401: obtain up to 480 (48 values of $\theta_{0}$ times up to 10 values of $\xi_{i}$)
402: statistically independent measurements per dec.\ band. Every
403: available pixel (histogram bin) pair in the
404: 48 hour angle vs.\ dec.\ histograms (Fig.\ \ref{fig3}) is used once and only
405: once. These one-dimensional histograms are assembled into a two-dimensional
406: histogram of FB vs.\ $\theta_{0}$ and $\xi_{i}$. This histogram contains
407: all our information about
408: FB and its statistical error. The fit of Eq.\ \ref{fitfun} to this 2-D
409: histogram (a simultaneous fit over $\theta_{0}$ and $\xi_{i}$) gives the
410: experimental values and errors of the Fourier coefficients for the harmonic
411: model of the anisotropy in a dec.\ band (Eq. \ref{skymod}). An example of the
412: 2-D histogram that is
413: being fit, and the result of the fit in $\theta_{0}$ for a single slice
414: in $\xi_{i}$ are shown in Fig.\ \ref{fig4}(a) and (b). The coherent signal in
415: FB is clearly seen in both. Figure \ref{fig4}(b) also shows the resulting
416: measured anisotropy reconstructed using the six fitted Fourier coefficients in
417: this example.
418:
419: The determination of the statistical errors on the Fourier coefficients
420: is straight-forward. The statistical errors on the number
421: of events ($N_{F}$ and $N_{B}$) are simply their square roots and these
422: errors are propagated to the error on FB:
423:
424: \begin{equation}
425: \sigma_{FB} = \sqrt{\frac{1}{N_{F} + N_{B}} (1 + FB^{2})} \approx \sqrt{\frac{1}{N_{F} + N_{B}}}
426: \end{equation}
427:
428: to a superb approximation. Since each FB bin in the 2-D histogram is
429: statistically independent of the others, the fit is that of a 2-D experimental
430: histogram with independent errors to the parameters of the theoretical
431: function of Eq.\ (\ref{fitfun}). This fit (using MINUIT \citep{MIN}) propagates the
432: errors on the experimental data into errors on the parameters of the theory,
433: namely the six Fourier coefficients of the anisotropy.
434:
435: It is noted that the simultaneous fit of the different $\xi_{i}$ bands is
436: in effect an averaging over $\xi_{i}$ which means that we are also averaging
437: over any difference of energy in the data due to the dependence of atmospheric
438: depth on $\xi$. This averaging is well justified after the fact by
439: an examination of the $\xi$ dependence discussed in Section \ref{stdataerr}.
440:
441: Finally, we emphasize that the role of the forward-backward asymmetry
442: is over once the fit is done. The coefficients obtained are for the
443: anisotropy of the sky, defined as the fractional difference of the rate
444: at a given r.a.\ from its average over all r.a., for each dec.\
445: band. In Section \ref{results}, we tabulate the fit results for these
446: 18 dec.\ bands, as well as graphically display their behavior.
447:
448:
449:
450:
451: \subsection{Monte Carlo Tests of the Analysis Method}
452:
453: Tests of the analysis method were conducted using a Monte Carlo simulation that
454: takes as input a 2-D map containing any desired anisotropy in universal
455: time (UT) or sidereal time (ST) and then outputs events which
456: can then be analyzed using the method discussed in the previous section.
457:
458: We describe one important case in detail: a fixed UT signal will average
459: to zero in ST or AST, but a seasonally modulated day-night
460: anisotropy in UT will generate a false ``sideband" signal in ST and
461: AST. We use a time varying input in UT, where the magnitude of the UT
462: signal varies sinusoidally between zero and its maximum with a period of one
463: year, simulating an extreme seasonal variation. Figure \ref{fig5} shows the
464: results of an all-dec-band analysis in three time frames for this
465: test. The ST and AST frames, which have no input signal in the
466: MC, show a clear signal induced by the time varying UT input. In this example
467: the induced signal is at about half the amplitude of the UT time
468: variation. The phases of the signals in ST and AST are different but their
469: magnitudes are the same. Using this result we can estimate the systematic
470: errors in the sidereal signal by
471: examining the anti-sidereal signal. Since there are no physical processes
472: which occur in the anti-sidereal time frame the signal should be zero when
473: data sets of an integral number of years are used as seen from the Monte Carlo
474: checks above. Therefore a signal present in the non-physical AST frame is
475: interpreted as being induced by variations in time of the UT signal and such a
476: signal will also appear in the ST frame.
477:
478: Two other sets of MC tests were successfully passed by the FB method. The
479: first set tests the stability of the
480: analysis method by using various simple signals (e.g.\ isotropic sky or the
481: result of the analysis of real data) as input in ST with no input in UT or anti-sidereal
482: time. Analysis of the output of these tests
483: is consistent with the input within statistical errors. The UT and
484: anti-sidereal time (AST) maps in these tests are consistent with isotropic
485: as long as an integral number of years is used. This is to be expected
486: since a static point in ST moves across the UT frame and returns to
487: its original position after one year. In AST this point will move
488: across two times in one year. Therefore when the declination band is
489: normalized a sidereal signal will average to zero in both UT and AST.
490: This cancellation is of course also true for the sidereal sky
491: with a signal fixed in UT (e.g.\ day, night effects).
492:
493: The second set of tests shows that no false harmonic features are
494: induced from isolated structures (e.g.\ a square well deficit centered around
495: $\sim185^{\circ}$ or the known Galactic equator excess of
496: gamma-rays \citep{GalRidge1,GalRidge2,GalRidge3}). These tests show that no extra signals are
497: induced except in the cases of a sharply discontinuous input and a time
498: varying UT input. The discontinuous input induces features due to the use of
499: only three harmonics as is expected from Fourier theory.
500:
501:
502:
503: \section{Results}
504:
505:
506: \subsection{Multi-dec-band Results and the Sidereal Sky Map}
507: \label{results}
508:
509: Table I is a summary of the harmonic fit parameters for the sidereal sky
510: fractional anisotropy obtained in each of 18 individual dec bands. The
511: $\chi^2$ for the fit and the sample size in each dec.\ band is also listed
512: in this table. Three harmonics were chosen as they give a better
513: $\chi^2$/ndf than one or two harmonics whereas there was little improvement
514: in using four. Fig. \ref{fig6} displays the anisotropy profiles in RA for
515: eight adjacent dec.\ bands in the table. As seen from the definition of the
516: anisotropy in Eq. \ref{skymod}, the profiles show the deviation from their
517: average over all r.a., so that the area below and above the reference level
518: of zero is the same: the existence of a
519: deficit in some region implies an excess elsewhere.
520:
521: Figure \ref{fig7} assembles all 18 anisotropy profiles in their
522: respective dec.\ positions into a 2-dimensional color display of the local
523: anisotropy, with red representing an excess and blue representing a
524: deficit. Care must be taken in interpreting Fig.\ \ref{fig7}. As
525: explained before, our measurement is in the r.a.\ direction only, so that this
526: picture does not purport to be the complete anisotropy of the celestial
527: sphere; nor can the complete anisotropy be inferred from our
528: results. Fig.\ \ref{fig7} is a two-dimensional display of one-dimensional
529: information, the fractional
530: anisotropy in r.a. What can be fairly compared across the dec bands
531: are the shape and strength of the fractional anisotropy in the r.a.\
532: direction; there is no information on the
533: r.a.\ averaged cosmic ray rate difference from one dec.\ band to
534: another, i.e.\ the anisotropy in the dec.\ direction.
535:
536:
537: The data and their analysis are completely independent for each dec.\
538: band, but a striking commonality of the anisotropy behavior is seen in
539: both Figs. \ref{fig6} and \ref{fig7}. Adjacent dec.\ bands have very
540: similar profiles, in shape and in strength. The conclusion is that there is
541: indeed a coherent anisotropy signal over a large swath of the sky. The
542: dominant feature is a prominent valley or deficit region extending from
543: $150^{\circ}$ to $225^{\circ}$ in r.a., and clearly visible in all dec.\ bands
544: between $-10^{\circ}$ and $45^{\circ}$. The decrease of the depth of this
545: valley, towards large values of dec., seen in the color picture of
546: Fig.\ \ref{fig7}, is explained at least in part by the fact that the
547: circle in the celestial sphere generated by the earth's rotation gets smaller
548: and smaller, shrinking to a single point as dec.\ approaches $90^{\circ}$. The
549: r.a.\ position of the valley minimum appears dec.\ independent in
550: Figs.\ \ref{fig6} and \ref{fig7}. Figure \ref{fig8} plots the
551: amplitude and phase of the lowest harmonic versus dec., confirming both
552: these trends; the amplitude is seen to decrease as dec.\ gets larger exhibiting
553: the expected cos(dec) dependence, while
554: the phase is quite stable for dec.\ below $60^{\circ}$, where the
555: amplitude is large enough for the phase to be well defined.
556:
557:
558: \subsection{Single-dec-band Results for ST, UT, and AST Time Frames}
559: \label{sbresults}
560:
561: As a complement to the multi-dec-band analysis of section 4.1
562: we do a single-band analysis by projecting the hour angle vs.\ dec.\ histograms onto
563: the horizontal axis of Fig.\ \ref{fig3}, thus integrating all dec.\ bands, then
564: performing the same anisotropy analysis on this single band, to
565: obtain the overall r.a.\ behavior. Examining the resulting single band
566: anisotropy, with only 6 fourier coefficients, gives a convenient and
567: economical way to compare and contrast the behavior in all three coordinate
568: systems corresponding to sidereal, universal and anti-sidereal time. Table 2
569: gives the fitted harmonic fit coefficients for ST, UT, and AST, and
570: Fig.\ \ref{fig9} displays the resulting profiles for all three
571: coordinate systems.
572:
573: The sidereal-frame profile of Fig.\ \ref{fig9} clearly reproduces the
574: dominant feature seen in the multiband analysis, the valley centered at
575: $189^{\circ}$ in r.a. The single band valley depth (SBVD) is an alternate
576: measure of the strength of the sidereal anisotropy and is found to be SBVD =
577: (2.49 $\pm$ 0.02 stat.\ $\pm$ 0.09 sys.) $\times 10^{-3}$, a
578: $22.6\sigma$ signal. The method of systematic error estimation will be discussed
579: in Section \ref{stdataerr}.
580:
581: The Compton Getting effect due to the earth's motion around the sun gives
582: a predicted anisotropy in universal-time frame which is a dipole
583: (the lowest harmonic), plotted in Fig. \ref{fig9} together with
584: our experimental result. Because the predicted signal amplitude is a factor
585: of five smaller than the observed sidereal time signal, and is
586: correspondingly more susceptible to systematic distortions, this is a
587: stringent test of the FB method. The experimental anisotropy profile is
588: indeed dominated by the lowest harmonic as expected, has the predicted
589: amplitude and reasonably close, though not in agreement (within errors) with
590: the predicted phase. A detailed analysis of
591: the UT measurement, including systematic errors, is deferred
592: to a future publication. We consider our observation of
593: the Earth-motion Compton-Getting anisotropy signal in UT a powerful check
594: of the analysis method, and thus a verification of the reliability of the
595: result for the sidereal sky.
596:
597:
598:
599: Since the anti-sidereal reference frame of Fig. \ref{fig9} is
600: non-physical, we expect no real signal there, while systematic effect
601: distortions will show up in this frame also. Indeed this signal, plotted in Fig.\ \ref{fig9}, is seen to be a factor
602: of 3 smaller than the UT signal and thus a factor of 15 smaller than the
603: ST signal. The observed AST signal will be used in the estimation of systematic
604: errors in the next section.
605:
606:
607:
608:
609:
610:
611:
612:
613: \subsection{Stability of Data and Systematic Errors}
614: \label{stdataerr}
615:
616: As a verification of the threshold independence and overall detector rate
617: independence of the method and results, we raise the PMT multiplicity
618: (the number of PMTs hit during an event in the air shower layer of the
619: detector) requirement in the trigger from its hardware value of around 75
620: in several steps up to a value of about 400. Figure \ref{fig10} shows
621: the number of events and the fractional
622: anisotropy valley depth SBVD as a function of the PMT multiplicity. The valley
623: depth is seen to be stable over a factor of five in trigger threshold
624: corresponding to a factor of $\sim 25$ in trigger rate. This is a direct
625: confirmation that the result is indeed impervious to the detector rate and
626: threshold, as expected from the rate cancellation of the FB method, explained
627: in Section \ref{fbahm}.
628:
629: In section 3.1 it is mentioned that the analysis method averages over different values of
630: $\xi$, from $2.5^{\circ}$ to $47.5^{\circ}$. Figure \ref{fig11} compares
631: the single-band profiles when the full data set is split in two and
632: analyzed separately, where one set contains data with $\xi$ ranging
633: from $0^{\circ}$ to $25^{\circ}$ and the other $25^{\circ}$
634: to $50^{\circ}$. The good agreement of these two profiles demonstrates that
635: the procedure of simultaneously fitting over all $\xi$ is well justified.
636:
637:
638: To check for the possibility of a dependence on seasons, we break the data
639: into three seasonal sets, defined as: Apr.-Jul., Jul.-Nov., and
640: Nov.-Apr.\ corresponding to average local weather periods of: warm with low
641: precipitation, high precipitation, and ice and snow respectively. No
642: significant changes are seen in Fig.\ \ref{fig11} which shows the three
643: single-band profiles superimposed.
644:
645:
646: To test whether the observed ST signal is a sidereal phenomena
647: and not due to universal time effects, such as weather, we break the data into two
648: month periods. On these short time scales there will not be a cancellation
649: between the ST, UT and AST time frames since a fixed position in ST
650: corresponds to a nearly constant position in UT. Fig.\ \ref{fig12} shows the position of
651: the observed minima in UT for each two month period over seven years. The solid
652: line shows the expected position of a fixed sidereal minimum located at $189^{\circ}$ r.a. If
653: the minimum observed at $189^{\circ}$ r.a. is a dominant feature of the
654: sky, then the minimum in UT should follow the sawtooth pattern
655: of the solid lines in the figure. If, on the other hand, there were dominant
656: features in the UT sky, the positions of the minima should not correlate at all
657: with this sawtooth pattern. The observed correlation between the observed and
658: predicted positions of the minima in Figure \ref{fig12} is a strong indication that
659: the ST signal is the dominant feature in UT on this time scale. The same
660: conclusion is reached when the corresponding analysis is carried out in the
661: AST time frame where the sawtooth pattern has double the frequency.
662:
663: To estimate systematic errors in ST or UT, for data sets
664: of a full year (or several full years) of data, we use the
665: anisotropy profiles in AST. The phase in the nonphysical
666: anti-sidereal frame sweeps continuously between $0^{\circ}$ and $360^{\circ}$
667: in ST or UT in the course of a year, so that any coherent signal is lost
668: because its phase is in effect randomized by this sweep. Since the
669: physical signals have been washed out, what is left are any systematic
670: distortions. We illustrate the logic and procedure
671: for the case of the AST profile seen in Fig.\ \ref{fig9} for the single band
672: analysis. While the deviation of this profile from zero gives the possible
673: magnitudes of the distortion of the valley depth, the phase relative
674: to the signal in ST is unknown. If a deficit region of distortion in the
675: AST profile were to coincide with the valley
676: it would increase the valley depth SBVD, an excess would decrease
677: it, and a zero crossing would leave it unchanged. We account
678: for all these possibilities by projecting the AST curve of Fig.\ \ref{fig9} onto the vertical axis, to obtain the histogram shown in Fig.\ \ref{fig13} of all the
679: possible distortions of SBVD that can result from this profile. We then
680: take the RMS deviation of this histogram, $\sigma_{RMS} = 0.09 \times 10^{-3}$
681: as the systematic error on the valley depth SBVD in the ST analysis.
682:
683:
684:
685:
686:
687: \subsection{Time Evolution of the Sidereal Anisotropy}
688: \label{timeev}
689:
690: To examine any time evolution of the anisotropy the data were split into
691: seven year-long sets from July 2000 to July 2007. Figure \ref{fig14} shows the
692: three amplitudes and three phases obtained by fitting data from each yearly
693: set using the single band analysis (over all dec.). There is a stability
694: in phase of all three harmonics over these seven years and in amplitude of the
695: two higher harmonics. The fundamental harmonic, however, shows a clear increase in
696: amplitude with time.
697:
698: In order to quantify this increase in a
699: precise manner, we look at the time evolution of the single band valley depth
700: (SBVD) in the all dec.\ analysis, defined in
701: Sec.\ \ref{sbresults} as a single-parameter measure of the strength of
702: the anisotropy. Using this parameter allows us to include all three
703: harmonics as well as utilize the systematic error estimation procedure
704: outlined in the previous section. First we confirm that the valley
705: is stable in position over time. The position in r.a.\ of the minima
706: over the course of seven years shows very little variation as is expected
707: from the stability of the harmonic phases seen in Fig.\ \ref{fig14}. Fitting
708: the position of the seven yearly minima to a constant yields a value
709: of $189^{\circ} \pm 1^{\circ}$ r.a.\ with a $\chi^{2}/ndf = 4.5/6$. This
710: lack of change in position over time is what one would expect from an actual
711: sidereal signal.
712:
713: Figure \ref{fig15}, SBVD vs.\ year, shows that there is strong
714: evidence of a strengthening of the valley depth over the seven year
715: span of this data set.
716:
717: To test the robustness of this time dependence a number of checks were
718: done. As a test that is completely different from the insensitivity of
719: anisotropy strength to trigger
720: thresholds (described in Section \ref{stdataerr} and Fig.\ \ref{fig10}), we
721: have done a direct check of whether the time-dependence of SBVD itself is
722: threshold dependent. For several raised multiplicity thresholds between 90
723: and 280 PMTs hit in the top layer of the pond, the same time dependence is seen; the
724: yearly trend does not disappear.
725:
726: To see that this is a sidereal effect and not a detector
727: effect we look at the yearly time evolution for the universal time and
728: anti-sidereal time signals. Figure \ref{fig16} shows the amplitudes of the
729: three fit parameters for the single band analysis (all dec.) in both UT and
730: AST; the earth motion Compton-Getting effect in UT should have no
731: time dependence of the amplitude. The amplitudes of the harmonics in UT
732: are constant over this seven year data set, within the errors, as well
733: as their phases (not shown in the figure). With respect to the amplitudes of
734: the harmonics in AST, these appear to be significantly larger in some
735: years, but even the largest amplitudes are 5 to 10 times less than those
736: seen in ST. From these tests it thus appears that time dependent
737: detector effects cannot account for the observed strong time dependence of the
738: sidereal anisotropy.
739:
740:
741: \section{Conclusions}
742:
743:
744: Previous experiments such as the Tibet Air Shower Array, with a modal energy of
745: 3 TeV, and Super-Kamiokande-I, with a median energy of 10 TeV, have identified two coincident regions of
746: interest in their sidereal observations: an excess located at
747: $\sim75^{\circ}$ r.a.\ or ``tail-in" anisotropy, and a deficit at
748: $\sim200^{\circ}$ r.a.\ or ``loss-cone" anisotropy \citep{TB,SK}.
749: Both of these regions are consistent with Milagro observations.
750: The ``loss-cone" coincides with the deep central-deficit region seen in
751: this analysis while the narrow ``tail-in" excess is clearly observed in another Milagro analysis which is
752: sensitive to features with extent smaller than $\sim30^{\circ}$ in r.a. (see \citet{GW} region A).
753:
754: The strengthening of the signal in the central-deficit region over time
755: is a result unique to this analysis. Tibet found no evidence of time variation
756: comparing data split into two five-year periods, 1997-2001 and
757: 2001-2005. However, only the second of these time periods overlaps with our data set for which
758: the average value we observe in the deficit region agrees
759: well with their measured deficit.
760:
761:
762: The anisotropy observed in the galactic cosmic rays could arise from a number of
763: possible effects. The Compton-Getting effect (CG) predicts that due to the
764: motion of the solar system around the galactic center through the rest frame
765: of the cosmic-ray plasma an anisotropy is induced with the form of a dipole
766: with a maximum in the direction of motion. For no co-rotation of the cosmic
767: ray plasma with the Galaxy, the magnitude
768: of the anisotropy is calculated to be $0.35\%$ given our speed of
769: $\sim220 km s^{-1}$, while at the other extreme of full co-rotation, the
770: anisotropy would be zero. No evidence of a Galactic CG anisotropy was seen in
771: \cite{TB}. For no co-rotation, the dipole should have a maximum at r.a.\ = $315^{\circ}$
772: and dec.\ = $48^{\circ}$, and a minimum at r.a.\ = $135^{\circ}$ and
773: dec.\ = $-48^{\circ}$. Since our analysis method yields only a projection of the anisotropy the observed CG effect will be slightly different from the true
774: effect. The CG effect we expect to see
775: in this analysis was determined from Monte Carlo simulation and found to be
776: a dipole with a maximum of $0.14\%$ at $315^{\circ}$ r.a.\ for the declination
777: range between $50^{\circ}$ to $60^{\circ}$. This range was considered to try
778: to reduce effects from the central-deficit region. For the actual data,
779: fitting a single harmonic to the projection in r.a.\ corresponding
780: to declinations $50^{\circ}$ to $60^{\circ}$ gives a $\chi^{2}/d.o.f.\ = 11505/998$
781: which is clearly a poor fit. Although this suggests that the observed
782: anisotropy is not dominated by the galactic
783: Compton-Getting effect, its contribution to the anisotropy cannot be
784: ruled out.
785:
786:
787: In addition to the Compton-Getting effect there is expected to be
788: an anisotropy stemming from the diffusion of cosmic-rays in the
789: interstellar medium. At high energies the main effects are
790: expected to be mainly due to the distribution of discrete CR sources and the
791: structure of the galactic magnetic field in the
792: neighborhood of the solar system. One study conducted consists of a simple
793: diffusion model assuming increased
794: production in the galactic disk due to
795: supernova remnants (SNR) \citep{PtSNR,SMP}. Also considered was
796: the diffusion of CR out of the galactic halo. This is attractive since we
797: have a deficit region in the general direction of the
798: North Galactic Pole which could be due in part to this diffusion. The main
799: contribution of CR from
800: SNR was considered for sources with
801: distances from Earth of $<1$ kpc and
802: ages $<0.05$ Myr. Calculations performed \citep{PtSNR,SMP} using these
803: sources and taking into account CR re-acceleration as well as diffusion out of
804: the galaxy gives an anisotropy
805: about 3 times greater than observed, with the main source of the
806: anisotropy due to the Vela SNR located at $128^{\circ}$ r.a.\ and
807: $-45.75^{\circ}$ dec. This model only predicts the magnitude of the expected
808: anisotropy, not its exact phase. It also is remarked that this
809: is a simplified model which assumes an isotropic diffusion tensor which is not
810: explicitly known to be true at these energies.
811:
812:
813: At energies of $\sim1$ TeV the heliosphere is
814: believed to have an influence on the distribution of CR \citep{NgLC,NgTI}. One
815: possible reason for the modulation of the anisotropy on the observed time scale
816: could be due to variations in the heliosphere since we know that it changes
817: in relation to solar output. It is noted
818: that our data begins at the solar maximum and ends near the solar minimum.
819: A recent derivation of the diffusion tensor contains a
820: new component due to perpendicular spatial diffusion which is expected to be an
821: important factor in understanding the anisotropy due to the Galactic disk
822: as well as the modulation of CR in the outer heliosphere \citep{PerpD}. Finding
823: a consistent explanation of the observed anisotropy
824: and especially its time evolution will be a challenge.
825:
826:
827:
828: \section{Acknowledgments}
829: We acknowledge Scott Delay and Michael Schneider for their dedicated efforts in the construction and maintenance of the Milagro experiment. This work has been supported by the National Science Foundation (under grants PHY-0245234, -0302000, -0400424, -0504201, -0601080, and ATM-0002744) the US Department of Energy (Office of High-Energy Physics and Office of Nuclear Physics), Los Alamos National Laboratory, the University of California, and the Institute of Geophysics and Planetary Physics.
830:
831:
832:
833: \clearpage
834:
835: \begin{thebibliography}{}
836:
837:
838:
839: \bibitem[Abdo et al.(2007)]{GalRidge2} Abdo, A. A., et al. 2007, \apj, 664, L91
840:
841: \bibitem[Abdo et al.(2008)]{GalRidge3} Abdo, A. A., et al. 2008, \apj, 688, 1078
842:
843: \bibitem[Abdo et al.(2008)]{GW} Abdo, A. A., et al. 2008, Phys. Rev. Lett., 101, 221101
844:
845: \bibitem[Aglietta et al.(1996)]{ET} Aglietta, M., et al. 1996, \apj, 470, 501
846:
847: \bibitem[Agostinelli et al.(2003)]{GEANT} Agostinelli, S., et al. 2003, Nucl. Instrum. Meth., A506, 250
848:
849: \bibitem[Amenomori et al.(2004)]{TBCG} Amenomori, M., et al. 2004, Phys. Rev. Lett., 93, 61101
850: \bibitem[Amenomori et al.(2006)]{TB} Amenomori, M., et al. 2006, Science, 314, 439
851: \bibitem[Atkins et al.(2004)]{Milagro} Atkins, R., et al. 2004 \apj, 608, 680
852: \bibitem[Atkins et al.(2005)]{GalRidge1} Atkins, R., et al. 2005 Phys. Rev. Lett., 95, 251103
853: \bibitem[Band et al.(1989)]{FBCA} Band, H. R., et al. 1989 Phys. Lett. B, 218, 369
854: \bibitem[Berezinskii et al.(1990)]{AsCR} Berezinskii, V. S., Bulanov, S. V., Dogiel, V. A., \& Ptuskin, V. S. 1990, Astrophysics of Cosmic Rays (Amsterdam: North-Holland)
855: \bibitem[Compton \& Getting(1935)]{CG} Compton, A. H., \& Getting, I. A. 1935, Phys. Rev., 47, 817
856: \bibitem[Farley \& Storey(1954)]{AST} Farley, F., \& Storey, J. 1954, Proc. Phys. Soc. A, 67, 996
857: \bibitem[Guillian et al.(2007)]{SK} Guillian, G., et al. 2007, Phys. Rev. D, 75
858:
859: \bibitem[Heck et al.(1998)]{CSKA} Heck, D., Knapp, J., Capdevielle, J. N., Schatz, G., \& Thouw, T. 1998, Forschungszentrum Karlsruhe Report FZKA, 6019
860:
861: \bibitem[James(2000)]{MIN} James, F. 2000, MINUIT - Function Minimization and Error Analysis - Reference Manual (Geneva, CH: CERN), \\
862: \url{http://wwwasdoc.web.cern.ch/wwwasdoc/minuit/minmain.html}
863:
864: \bibitem[Nagashima et al.(1989)]{NgLC} Nagashima, K., et al. 1989, Nuovo Cimento C, 12 695
865: \bibitem[Nagashima et al.(1998)]{NgTI} Nagashima, K., Fujimoto, K., \& Jacklyn, R. M. 1998, JGR, 103, 17429
866:
867: \bibitem[Panov et al.(2006)]{ATIC} Panov, A. D., et al. 2006, Bull. Russ. Acad. Sci., Phys. in press (astro-ph/0612377v1)
868:
869: \bibitem[Ptuskin et al.(2006)]{PtSNR} Ptuskin, V. S. Jones, F. C. Seo, E. S. \& Sina, R. 2006, AdSpR, 37, 1909
870: \bibitem[Schlickeiser et al.(2007)]{PerpD} Schlickeiser, R., Dohle, U., Tautz, R. C., \& Shalchi, A. 2007, ApJ, 661, 185
871: \bibitem[Strong et al.(2007)]{SMP} Strong, A., Moskalenko, I., \& Ptuskin, V. 2007, ARN\&PS, 57, 285
872: %\bibitem[Yodh et al.(2007)]{BAGY} Yodh, G.B., et al. 2007, Proceedings of the 30th ICRC, Merida, Mexico
873:
874:
875: \end{thebibliography}
876:
877:
878:
879:
880:
881:
882: \clearpage
883:
884: \begin{figure}
885: \epsscale{.80}
886: \plotone{f1}
887: \caption{Diagram showing the definition of $\xi$ used in the calculation of the
888: forward-backward asymmetry for a single dec.\ band and a given 30 minute
889: histogram. $\xi$ is in the direction of hour angle. See text in Section 3.1 for
890: details.}\label{fig1}
891: \end{figure}
892:
893: \clearpage
894:
895: \begin{figure}
896: \epsscale{.80}
897: \plotone{f2}
898: \caption{Representative plot of the number of events collected vs.
899: azimuthal angle for the Milagro detector over the course of three days. The
900: observed variation is the inherent geometric asymmetry in the detector which is
901: at the level of $\sim 10 \%$.}\label{fig2}
902: \end{figure}
903:
904:
905:
906: \clearpage
907:
908: \begin{figure}
909: \epsscale{0.80}
910: \plotone{f3}
911: \caption{Example histogram showing the number of cosmic-ray events vs.\ hour
912: angle and declination containing seven years of data for a single sidereal \
913: half-hour. The solid circles are an example pair of pixels shown in Fig. 1
914: (with $\xi = 42.5^{\circ}$) used in the calculation of the FB given by
915: Eq.(1) \& Eq.(3).}\label{fig3}
916: \end{figure}
917:
918:
919:
920:
921:
922: \clearpage
923:
924: \begin{figure}
925: %\epsscale{.80}
926: \centering
927:
928: \begin{tabular}{ccc}
929:
930: & \plotone{f4a} & \\
931: & \plotone{f4b} & \\
932:
933: \end{tabular}
934:
935:
936: %\plotone{FB_Alphapro_Ani}
937:
938: \caption{a) Sample 2-D histogram of FB vs. $\xi$ and r.a.\ for a single dec.\
939: band ($35^{\circ} - 40^{\circ}$) calculated using Eq.\ \ref{eqn1} for seven years
940: worth of data (the binning reflects that the data is
941: collected over half hour intervals which are parameterized by $\theta_{0}$). b) Sample histogram showing
942: the result of the 2-D fit of Eq.\ \ref{fitfun} to a) for a single slice
943: in $\xi = 40^{\circ} - 45^{\circ}$ (in black, left axis), and the reconstructed
944: anisotropy of Eq.\ \ref{skymod} using the six Fourier coefficients obtained by the 2-D fit to a) (in red, right axis).}\label{fig4}
945: \end{figure}
946:
947:
948: \clearpage
949:
950: \begin{figure}
951: \plotone{f5}
952: \caption{Results of an all-dec-band anisotropy analysis using a MC
953: generated UT signal. The signal consists of square well deficit
954: from $150^{\circ}$ to $210^{\circ}$ in UT with an amplitude varying
955: sinusoidally in time from 0.000 to 0.003. The signal seen in the ST
956: and AST frames is induced by the UT time variation. The width of the curves
957: reflects the statistical error.}
958: \label{fig5}
959: \end{figure}
960:
961:
962: \clearpage
963:
964: \begin{figure}
965: \plotone{f6}
966: \caption{Profiles in r.a.\ for individual $5^{\circ}$ dec.\ bands
967: from $10^{\circ}$ to $50^{\circ}$ used in the 2-D map seen in
968: Figure \ref{fig7}. The width of the lines reflects the statistical error.}
969: \label{fig6}
970: \end{figure}
971:
972:
973: \clearpage
974:
975:
976:
977:
978: \begin{figure}
979: \epsscale{0.80}
980: \plotone{f7}
981: \caption{Result of a harmonic fit to the fractional difference of the
982: cosmic-ray rates from isotropic in equatorial coordinates as viewed by
983: Milagro for the years 2000-2007. The color bin width is $1.0\times10^{-4}$
984: reflecting the average statistical error. The two black lines show the position
985: of the Galactic Equator and the solid circle shows the position of the
986: Galactic North Pole. This map is constructed by combining 18 individual
987: profiles of the anisotropy projection in r.a.\ of width $5^{\circ}$ in
988: dec. It is not a 2-dimensional map of the sky. The median energy of the events in this map is 6 TeV.}
989: \label{fig7}
990: \end{figure}
991:
992: \clearpage
993:
994: \begin{figure}
995: \plotone{f8}
996: \caption{Amplitude and phase of the fundamental harmonic obtained from a fit
997: to seven years of data for each $5^{\circ}$ declination slice. The error bars
998: are statistical.}\label{fig8}
999: \end{figure}
1000:
1001: \clearpage
1002:
1003:
1004:
1005: \begin{figure}
1006: \plotone{f9}
1007: \caption{Anisotropy constructed by fitting a single projection containing
1008: data collected from all declinations collected over seven years in ST (red,
1009: left axis), UT (black, right axis) and AST (blue, right axis). The right axis
1010: (UT \& AST) is expanded by a factor of 4 compared to the left axis (ST). The
1011: dashed curve is
1012: the predicted Compton-Getting (C-G) effect due to the
1013: Earth's motion around the sun which has an expected maximum value at 6 hr local solar
1014: time ($196.3^{\circ}$ UT given the longitude of Milagro). In these plots the
1015: width of the curve reflects the statistical error.}\label{fig9}
1016: \end{figure}
1017:
1018:
1019:
1020: \clearpage
1021:
1022: \begin{figure}
1023: %\epsscale{0.80}
1024:
1025:
1026: \plotone{f10}
1027:
1028:
1029: \caption{Total number of events over a six year period (left axis) and valley
1030: depth SBVD (right axis) vs.\ PMT multiplicity. The PMT multiplicity is defined as being the number of PMTs hit
1031: in the top layer of the pond for a given event. The lack of change in SBVD
1032: seen compared to the large decrease in the number of triggers shows explicitly
1033: the insensitivity of the analysis to variations in the trigger threshold.}
1034: \label{fig10}
1035: \end{figure}
1036:
1037:
1038: \clearpage
1039:
1040: \begin{figure}
1041: \plotone{f11}
1042: \caption{Single-dec-band analysis showing seven years of data
1043: split into two sets: $\xi < 25^{\circ}$ (in yellow), and $25^{\circ} \leq \xi
1044: \leq 50^{\circ}$ (in green). Also plotted is the seven years of
1045: data split into three seasonal sets: Summer-Fall in black, Winter-Spring in
1046: blue, and Spring-Summer in red. The error bars are statistical +
1047: systematic. The independence of the analysis on $\xi$ as well as
1048: the lack of seasonal variation can be seen here. }
1049: \label{fig11}
1050: \end{figure}
1051:
1052: \clearpage
1053:
1054: \begin{figure}
1055: \epsscale{0.80}
1056: \plotone{f12}
1057: %\centering
1058: %\includegraphics[angle=270,width=0.9\textwidth]{UniMinPos}
1059: \caption{Position of minima in universal time vs.\ Modified Julian Date (MJD)
1060: for data collected over the 42 two-month periods starting in July 2000 and ending in July 2007. The error bars are the stat.\ errors. The solid line is the
1061: calculated position in universal time of a constant sidereal position at r.a.\
1062: equal to $189^{\circ}$. The excellent agreement between the data and the
1063: calculation shows that the UT signal is dominated by the ST signal.}
1064: \label{fig12}
1065: \end{figure}
1066:
1067: \clearpage
1068:
1069: \begin{figure}
1070: \epsscale{0.80}
1071: \plotone{f13}
1072: \caption{Histogram showing the projection of the AST curve from
1073: Figure \ref{fig9} onto the anisotropy axis. The RMS of this distribution
1074: is $0.09 \times 10^{-3} $ which is used to determine the systematic error on
1075: the sidereal signal.}\label{fig13}
1076: \end{figure}
1077:
1078:
1079: \clearpage
1080:
1081: \begin{figure}
1082: \centering
1083: %\epsscale{0.80}
1084: \begin{tabular}{ccc}
1085:
1086: & \plotone{f14a} & \\
1087: & \plotone{f14b} & \\
1088:
1089: \end{tabular}
1090: \caption{a) Sidereal time amplitudes of the three fit harmonics for the single band (all dec.) analysis. b) Sidereal time phases of the three fit harmonics for the single band analysis. Both plots contain seven yearly data sets from July 2000 to July 2007. The
1091: error bars are statistical.}
1092: \label{fig14}
1093: \end{figure}
1094:
1095: %\clearpage
1096: %
1097: %\begin{figure}
1098: %\epsscale{0.80}
1099: %\plotone{SB_MinPos}
1100: %\caption{Position of the minimum in r.a.\ using the all dec.\ analysis
1101: %for yearly sets from 2000-2007. The
1102: %error bars are the linear combination of the stat.\ \& sys.\ errors. The fit to
1103: %a constant value is $189^{\circ} \pm 1^{\circ}$. The $\chi^{2}$ / ndf is 4.5/6.}\label{fig8}
1104: %\end{figure}
1105:
1106:
1107:
1108: \clearpage
1109:
1110: \begin{figure}
1111: \epsscale{0.80}
1112: \plotone{f15}
1113: \caption{Valley depth in the all-dec-band analysis
1114: (SBVD) vs.\ MJD for yearly sets from July 2000 to July 2007. The error bars are the linear
1115: sum of the statistical \& systematic errors. The solid line is the fit
1116: to a constant value and the dashed is the linear two-parameter fit. The
1117: $\chi^{2}$/ndf for the fits are 86.2/6 and 4.4/5 respectively. The fit
1118: parameter in the flat case is $(2.39\pm0.08)\times10^{-3}$; the two fit
1119: parameters to the function $A(MJD)=p_0(MJD-53000)+p_1$ are:
1120: $p_0 = (0.97\pm0.11)\times10^{-6}$ and $p_1 = (2.34\pm0.08)\times10^{-3}$.}
1121: \label{fig15}
1122: \end{figure}
1123:
1124:
1125: \clearpage
1126:
1127: \begin{figure}
1128: \centering
1129: %\epsscale{0.80}
1130: \begin{tabular}{ccc}
1131:
1132: & \plotone{f16a} & \\
1133: & \plotone{f16b} & \\
1134:
1135: \end{tabular}
1136: \caption{a) Universal time fit amplitudes for the single band (all dec.) analysis for seven yearly data sets from July 2000 to July 2007. b) Anti-sidereal time fit amplitudes for the single
1137: band analysis for yearly data sets. For the UT fundamental harmonic only we
1138: show the statistical error + an estimate of the systematic error. For AST the
1139: error bars are only statistical. Note the
1140: lack of any definite trend, as opposed to what is seen in ST (Fig. \ref{fig14})}
1141: \label{fig16}
1142: \end{figure}
1143:
1144:
1145:
1146:
1147: \clearpage
1148:
1149:
1150: \begin{deluxetable}{ccccccccc}
1151: \tabletypesize{\footnotesize}
1152: \tablecaption{Fit parameters to the sidereal anisotropy for all 18 declination bands.}\label{tab1}
1153:
1154: \tablehead{
1155: \colhead{Dec.} & \multicolumn{2}{c}{Fund.\ Harm.} & \multicolumn{2}{c}{1st Harm.} & \multicolumn{2}{c}{2nd Harm.} & \colhead{$\chi^2$/d.o.f.} & \colhead{Number}\\
1156:
1157: \colhead{(mean)} & \colhead{Amplitude} & \colhead{Phase} & \colhead{Amplitude} & \colhead{Phase} & \colhead{Amplitude} & \colhead{Phase} & \colhead{} & \colhead{of events}\\
1158: \colhead{} & \colhead{($\times10^{-3}$)} & \colhead{(deg)} & \colhead{($\times10^{-3}$)} & \colhead{(deg)} & \colhead{($\times10^{-3}$)} & \colhead{(deg)} & \colhead{} & \colhead{($\times10^{9}$)}
1159: }
1160:
1161: \startdata
1162:
1163: 77.5&0.54$\pm$0.26&2$\pm$27&0.37$\pm$0.14&13$\pm$11&0.10$\pm$0.11&-39$\pm$19&262.57/234&0.65\\
1164: 72.5&0.73$\pm$0.14&22$\pm$11&0.19$\pm$0.08&-25$\pm$12&0.06$\pm$0.06&11$\pm$19&266.50/282&1.38\\
1165: 67.5&0.72$\pm$0.09&23$\pm$7&0.06$\pm$0.05&-24$\pm$26&0.01$\pm$0.04&28$\pm$116&308.67/330&2.39\\
1166: 62.5&0.83$\pm$0.07&19$\pm$5&0.12$\pm$0.04&-65$\pm$10&0.15$\pm$0.04&-2$\pm$4&355.61/330&3.63\\
1167: 57.5&0.99$\pm$0.06&6$\pm$3&0.12$\pm$0.03&-42$\pm$8&0.15$\pm$0.03&3$\pm$4&379.61/378&4.98\\
1168: 52.5&1.10$\pm$0.05&8$\pm$3&0.22$\pm$0.03&-60$\pm$4&0.17$\pm$0.03&6$\pm$3&406.42/378&6.31\\
1169: 47.5&1.31$\pm$0.04&8$\pm$2&0.33$\pm$0.03&-63$\pm$2&0.21$\pm$0.02&2$\pm$2&498.02/426&7.51\\
1170: 42.5&1.71$\pm$0.04&8$\pm$1&0.44$\pm$0.02&-68$\pm$2&0.26$\pm$0.02&1$\pm$2&475.85/426&8.46\\
1171: 37.5&1.95$\pm$0.04&6$\pm$1&0.45$\pm$0.02&-73$\pm$1&0.24$\pm$0.02&3$\pm$2&472.71/426&9.07\\
1172: 32.5&2.04$\pm$0.04&10$\pm$1&0.47$\pm$0.02&-76$\pm$1&0.20$\pm$0.02&1$\pm$2&520.47/426&9.28\\
1173: 27.5&2.17$\pm$0.04&9$\pm$1&0.53$\pm$0.02&-78$\pm$1&0.14$\pm$0.02&0$\pm$3&551.53/426&9.07\\
1174: 22.5&2.39$\pm$0.04&11$\pm$1&0.52$\pm$0.02&-81$\pm$1&0.12$\pm$0.02&-12$\pm$3&564.14/426&8.44\\
1175: 17.5&2.56$\pm$0.05&12$\pm$1&0.57$\pm$0.03&-81$\pm$1&0.10$\pm$0.02&-32$\pm$5&523.45/378&7.45\\
1176: 12.5&2.62$\pm$0.06&9$\pm$1&0.61$\pm$0.03&-85$\pm$2&0.05$\pm$0.03&-28$\pm$9&397.37/330&6.17\\
1177: 7.5&2.81$\pm$0.07&5$\pm$1&0.58$\pm$0.04&-80$\pm$2&0.08$\pm$0.03&-39$\pm$7&355.24/282&4.74\\
1178: 2.5&3.05$\pm$0.10&7$\pm$2&0.61$\pm$0.05&-80$\pm$3&0.14$\pm$0.04&54$\pm$5&280.00/234&3.31\\
1179: -2.5&2.96$\pm$0.15&8$\pm$3&0.59$\pm$0.08&-89$\pm$4&0.23$\pm$0.06&56$\pm$5&271.14/186&2.04\\
1180: -7.5&3.80$\pm$0.38&11$\pm$6&0.42$\pm$0.19&-85$\pm$13&0.14$\pm$0.13&52$\pm$18&118.02/90&1.04\\
1181:
1182: \enddata
1183: \tablecomments{All quoted errors are statistical and are used in the
1184: calculation of $\chi^2$.}
1185: \end{deluxetable}
1186:
1187:
1188: \clearpage
1189:
1190:
1191: \begin{deluxetable}{cccccccc}
1192: \tabletypesize{\footnotesize}
1193: \tablecaption{Fit parameters to the ST, UT and AST anisotropies using
1194: data collected over all declinations.}\label{tab2}
1195:
1196: \tablehead{
1197: \colhead{Time} & \multicolumn{2}{c}{Fund.\ Harm.} & \multicolumn{2}{c}{1st Harm.} & \multicolumn{2}{c}{2nd Harm.} & \colhead{$\chi^2$/d.o.f.}\\
1198:
1199: \colhead{Frame} & \colhead{Amplitude} & \colhead{Phase} & \colhead{Amplitude} &
1200: \colhead{Phase} & \colhead{Amplitude} & \colhead{Phase} & \colhead{}\\
1201: \colhead{} & \colhead{($\times10^{-3}$)} & \colhead{(deg)} &
1202: \colhead{($\times10^{-3}$)} & \colhead{(deg)} & \colhead{($\times10^{-3}$)} & \colhead{(deg)} & \colhead{}
1203: }
1204:
1205: \startdata
1206:
1207: Sidereal&1.994$\pm$0.012&9.1$\pm$0.4&0.400$\pm$0.007&104.3$\pm$0.5&0.118$\pm$0.006&-0.9$\pm$1.0&936.3/426\\
1208:
1209: Universal&0.365$\pm$0.012&217.4$\pm$1.9&0.043$\pm$0.007&168.7$\pm$4.8&0.029$\pm$0.006&93.2$\pm$4.3&520.9/426\\
1210:
1211: Anti-Sidereal&0.120$\pm$0.012&47.1$\pm$5.8&0.019$\pm$0.007&155.5$\pm$11.1&0.013$\pm$0.006&4.9$\pm$9.4&425.9/426\\
1212:
1213: \enddata
1214: \tablecomments{All quoted errors are statistical and are used in the
1215: calculation of $\chi^2$. For this single-dec-band analysis, the systematic errors exceed the statistical ones.}
1216: \end{deluxetable}
1217:
1218:
1219:
1220:
1221:
1222:
1223:
1224: \end{document}
1225: