1: \documentclass[12pt,preprint]{aastex}
2: %\documentclass[manuscript]{aastex}
3: %\usepackage{color}
4: %\usepackage{graphicx}
5: %\usepackage{natbib}
6: %\usepackage{epsfig}
7: \def\comma{,~}
8: \newcommand{\mygloss}[1]{\renewcommand{\baselinestretch}{1.}\marginpar{\tiny{#1}}}
9:
10: \begin{document}
11:
12: \title{A Measurement of the Spatial Distribution of Diffuse TeV Gamma Ray Emission from the Galactic Plane with Milagro}
13:
14: \author{
15: A.~A.~Abdo,\altaffilmark{1}
16: B.~Allen,\altaffilmark{2}
17: T.~Aune,\altaffilmark{3}
18: D.~Berley,\altaffilmark{4}
19: E.~Blaufuss,\altaffilmark{4}
20: S.~Casanova,\altaffilmark{5}
21: C.~Chen,\altaffilmark{6}
22: B.~L.~Dingus,\altaffilmark{7}
23: R.~W.~Ellsworth,\altaffilmark{8}
24: L.~Fleysher,\altaffilmark{9}
25: R.~Fleysher,\altaffilmark{9}
26: M.~M.~Gonzalez,\altaffilmark{10}
27: J.~A.~Goodman,\altaffilmark{4}
28: C.~M.~Hoffman,\altaffilmark{7}
29: P.~H.~H\"untemeyer,\altaffilmark{7}
30: B.~E.~Kolterman,\altaffilmark{9}
31: C.~P.~Lansdell,\altaffilmark{11}
32: J.~T.~Linnemann,\altaffilmark{12}
33: J.~E.~McEnery,\altaffilmark{13}
34: A.~I.~Mincer,\altaffilmark{9}
35: I.~V.~Moskalenko, \altaffilmark{14}
36: P.~Nemethy,\altaffilmark{9}
37: D.~Noyes,\altaffilmark{4}
38: T.~A.~Porter,\altaffilmark{3}
39: J.~Pretz,\altaffilmark{7}
40: J.~M.~Ryan,\altaffilmark{15}
41: P.~M.~Saz~Parkinson,\altaffilmark{3}
42: A.~Shoup,\altaffilmark{16}
43: G.~Sinnis,\altaffilmark{7}
44: A.~J.~Smith,\altaffilmark{4}
45: A.~W.~Strong, \altaffilmark{17}
46: G.~W.~Sullivan,\altaffilmark{4}
47: V.~Vasileiou,\altaffilmark{4}
48: G.~P.~Walker,\altaffilmark{7}
49: D.~A.~Williams,\altaffilmark{3}
50: and
51: G.~B.~Yodh\altaffilmark{6}}
52:
53: \altaffiltext{1}{ Naval Research Laboratory, Washington, DC.}
54: \altaffiltext{2}{ Harvard-Smithsonian Center for Astrophysics, Cambridge, MA.}
55: \altaffiltext{3}{ University of California, Santa Cruz, CA.}
56: \altaffiltext{4}{ University of Maryland, College Park, MD.}
57: \altaffiltext{5}{ Max-Planck-Institut f\"ur Kernphysik, Heidelberg, Germany.}
58: \altaffiltext{6}{ University of California, Irvine, CA.}
59: \altaffiltext{7}{ Los Alamos National Laboratory, Los Alamos, NM.}
60: \altaffiltext{8}{ George Mason University, Fairfax, VA.}
61: \altaffiltext{9}{ New York University, New York, NY.}
62: \altaffiltext{10}{ Instituto de Astronom\'{i}a, Universidad Nacional Aut\'{o}noma de M\'{e}xico, D.F., M\'{e}xico.}
63: \altaffiltext{11}{ Institute for Defense Analyses, Alexandria, VA.}
64: \altaffiltext{12}{Michigan State University, East Lansing, MI.}
65: \altaffiltext{13}{NASA Goddard Space Flight Center, Greenbelt, MD.}
66: \altaffiltext{14}{HEPL \& KIPAC, Stanford University, Stanford, CA.}
67: \altaffiltext{15}{University of New Hampshire, Durham, NH.}
68: \altaffiltext{16}{Ohio State University, Lima, OH.}
69: \altaffiltext{17}{Max-Planck-Institut f\"ur extraterrestrische Physik, Garching, Germany.}
70:
71: \begin{abstract}
72:
73: Diffuse $\gamma$-ray emission produced by the interaction of
74: cosmic-ray particles with matter and radiation in the Galaxy can be used to
75: probe the distribution of cosmic rays and their sources in different regions
76: of the Galaxy. With its large field of view and
77: long observation time, the Milagro Gamma Ray Observatory is an ideal
78: instrument for surveying large regions of the Northern Hemisphere sky
79: and for detecting diffuse $\gamma$-ray emission at very high energies.
80: Here, the spatial distribution and the flux of the diffuse $\gamma$-ray emission
81: in the TeV energy range with a median energy of 15 TeV
82: for Galactic longitudes between 30$^\circ$ and 110$^\circ$ and between
83: 136$^\circ$ and 216$^\circ$ and for Galactic latitudes between -10$^\circ$
84: and 10$^\circ$ are determined.
85: The measured fluxes are consistent with predictions of the GALPROP model everywhere
86: except for the Cygnus region ($l\in[65^\circ,85^\circ]$).
87: For the Cygnus region, the flux is twice the predicted value.
88: This excess can be explained by the presence of active cosmic ray sources
89: accelerating hadrons which interact with the local dense interstellar
90: medium and produce gamma rays through pion decay.
91:
92:
93: \end{abstract}
94:
95: \keywords{gamma rays: observations}
96:
97: \section{Introduction}
98:
99:
100: One hundred years after their discovery, the origin, acceleration, and
101: propagation of Galactic cosmic
102: rays is still unclear. Supernova remnants (SNRs) and pulsars are the preferred
103: candidate sources of Galactic cosmic rays. The detection of TeV
104: $\gamma$ rays and X-rays from the same locations within SNRs provides
105: strong evidence that electrons are accelerated in SNRs~\citep{2006A&A...449..223A}.
106: However, no compelling evidence for the acceleration of hadrons in
107: SNRs has yet been found.
108: The Galactic diffuse $\gamma$-ray emission originates from
109: the interactions of cosmic rays (hadrons and electrons) with
110: the matter and radiation fields in the Galaxy.
111: Cosmic ray hadrons interact with matter producing neutral pions which
112: in turn decay into $\gamma$ rays while cosmic-ray electrons produce TeV
113: $\gamma$ rays by inverse Compton (IC) scattering off the interstellar
114: radiation fields. Therefore, $\gamma$ rays can provide information about
115: the density and spectra of cosmic rays throughout the Galaxy.
116: Gamma rays above 10 TeV, produced by the highest energy particles
117: accelerated in SNR shocks, can be used to probe the acceleration processes in Galactic sources.
118: The spatial distribution of these TeV $\gamma$ rays
119: can be compared to model predictions of the diffuse $\gamma$-ray
120: production from $\pi^0$ decays and IC scattering and thus
121: the relative contribution of hadronic and leptonic mechanisms
122: can be investigated.
123:
124: EGRET data on the diffuse emission from the Galactic plane
125: \citep{1997ApJ...481..205H} show a harder $\gamma$-ray spectrum
126: from the inner Galaxy than predicted on the basis of the cosmic-ray
127: spectrum and intensity measured at Earth
128: \citep{1993ApJ...416..587B,2004A&A...422L..47S,2004ApJ...613..956S}.
129: Many explanations for this ``GeV excess'' have been proposed,
130: including a harder Galactic proton or electron spectrum
131: \citep{1997A&A...318..925G,2000A&A...362..937A,2004A&A...422L..47S,2004ApJ...613..956S,1997JPhG...23.1765P}
132: and the annihilation of dark matter particles~\citep{2005A&A...444...51D}.
133: Recent studies of the EGRET data have also investigated experimental uncertainties
134: associated with the determination of the Galactic diffuse emission and are discussed in \citet{2007NuPhS.173...44M}.
135: \citet{2008APh....29...25S} concluded that the EGRET
136: sensitivity above 1~GeV may have been overestimated, while
137: \citet{2007arXiv0706.0503B} concluded it may have been underestimated,
138: leaving the situation unclear.
139:
140: The first measurement of diffuse emission above 3.5 TeV from a large region
141: of the Galactic plane (Galactic longitudes ${40}^\circ<l<{100}^\circ$)
142: indicated the existence of a TeV
143: excess~\citep{2005PhRvL..95y1103A, 2007APh....27...10P, 2007arXiv0711.2753C}.
144: More recent measurements of the diffuse emission near 12 TeV from the Cygnus
145: region of the Galaxy~\citep{2007ApJ...658L..33A} also show an excess when compared to predictions of
146: GALPROP, a numerical model of cosmic-ray propagation in the
147: Galaxy.
148: Recently, HESS has detected very high-energy (VHE) diffuse emission from the Galactic Center Ridge,
149: that is correlated with giant molecular clouds.
150: The spectrum of the diffuse emission from the
151: Galactic Center Ridge is
152: significantly harder than the spectrum of the diffuse emission
153: predicted by assuming the local cosmic-ray spectrum~\citep{2006Natur.439..695A}.
154: These previous results from Milagro and HESS support the hypothesis that the
155: cosmic-ray flux is likely to vary throughout the Galaxy.
156:
157: The Milagro~\citep{2004ApJ...608..680A} experiment is a water-Cherenkov detector at an altitude of 2630~m.
158: It is composed of a central 60~m $\times$ 80~m
159: pond with a sparse 200~m $\times$ 200~m array of 175 ``outrigger'' tanks surrounding it.
160: The pond is instrumented with two layers of photomultiplier tubes. The top, ``air-shower'',
161: layer consists of 450 PMTs under 1.4~m of water while the bottom, ``muon'', layer has
162: 273 PMTs located 6~m below the surface. The air-shower layer allows the accurate measurement
163: of shower particle arrival times used for direction reconstruction and triggering. The greater
164: depth of the muon layer is used to detect penetrating muons and hadrons. The outrigger array,
165: added in 2003, improved the angular resolution of the detector from $\sim$0.75$^\circ$
166: to $\sim$0.45$^\circ$ by providing a longer lever
167: arm with which to reconstruct events. Milagro's large field of
168: view ($\sim$2 sr) and high duty cycle ($>90$\%) allow it to monitor the entire overhead sky
169: continuously, making it well-suited to measuring diffuse emission.
170:
171: Here, the Milagro measurement of the diffuse emission around 15 TeV
172: from a region of the Galactic plane of longitudes $l\in[30^\circ,110^\circ]$
173: and $l\in[136^\circ,216^\circ]$, and latitudes $b\in[-10^\circ,10^\circ]$
174: is presented.
175: The measured $\gamma$-ray flux and the latitudinal and longitudinal
176: profiles of the emission are reported and compared to
177: predictions of the
178: GALPROP model~\citep{2000ApJ...537..763S,2004A&A...422L..47S,2004ApJ...613..956S,2008arXiv0804.1774P}.
179: In GALPROP, first the propagation of cosmic rays in the Galaxy is modeled, and
180: then the $\gamma$-ray emissivities are calculated using the propagated spectra
181: of cosmic rays and the gas and radiation densities. The ``conventional'' model is
182: tuned to reproduce the local direct cosmic-ray measurements.
183: The ``optimized'' model has been designed to reproduce the EGRET data
184: by relaxing the restriction from the local cosmic-ray measurements.
185: In this version of the model, the proton spectrum is constrained
186: by the cosmic-ray antiproton measurements and the electron spectrum is
187: constrained using the EGRET data themselves.
188:
189: Below, the course of the analysis is described, followed by the presentation
190: of the results and a comparison with GALPROP predictions. In the last section, likely
191: interpretations of the observations are discussed.
192:
193:
194: \section{Analysis}
195:
196: The Milagro data, collected between July 2000 and November 2007, were analyzed
197: using the method described in~\citet{2007ApJ...658L..33A}.
198: Only events with a zenith angle less than $45^\circ$ are included, which corresponds
199: to declinations between $-7^\circ$ and $81^\circ$.
200: The event excess is calculated using the background estimation method described
201: in~\citet{2003ApJ...595..803A} with the modification that the events
202: are weighted by a factor dependent on the $\gamma$-hadron separation parameter
203: $A_{4}$~\citep{2007ApJ...658L..33A}.
204: Only declinations $<70^\circ$ are considered. This choice is governed by the fact that
205: for $\delta >70^\circ$ the Galactic equator turns parallel to the right ascension axis.
206: This causes the ratio of on- to off-time in the background
207: calculation~\citep{2003ApJ...595..803A} to become too big for signal bin sizes of 2 deg longitude by 4 deg
208: latitude which is the bin size that is used in the Galactic longitude flux profile. As a result,
209: the present analysis is insensitive for $\delta >70^\circ$ or $l\in[111^\circ,135^\circ]$.
210:
211: Within the region studied here, Milagro has previously detected four sources and four
212: source candidates~\citep{2007ApJ...658L..33A,2007ApJ...664L..91A}. The contribution
213: from these sources is taken into account by modeling each source
214: as a two-dimensional Gaussian plus a constant. The source location $(RA,\delta)$,
215: the amplitude and radial width of the Gaussian, and the constant are determined
216: using a $\chi^2$ minimization. The excess from each source is then calculated bin by bin
217: using the resulting Gaussian function and subtracted from the total excess in
218: the $0.1^\circ\times 0.1^\circ$ bin event excess map of the Galactic plane. The
219: resulting diffuse event excess is converted to a flux with a Monte Carlo simulation of
220: extensive air showers (CORSIKA,~\citet{corsika}) and of the Milagro detector
221: (GEANT4,~\citet{geant}).
222: The diffuse flux is calculated assuming a power law photon spectrum with
223: a differential spectral index $\alpha=-2.75$.
224: This spectral index was chosen to match the cosmic-ray spectrum in the
225: energy range of this analysis (around 10~TeV).
226: For a spectral index of -2.75 the median energy of detected events used in this analysis
227: is 15 TeV.
228:
229: Studies of possible sources for systematic errors have been performed.
230: The size of the fit region around the eight sources and source candidates
231: was varied.
232: The Gaussian fits to the event excesses were performed in boxes centered around the sources
233: of $4^\circ\times 4^\circ$, $6^\circ\times 6^\circ$, and $8^\circ\times 8^\circ$.
234: The flux determination was also repeated for spectral indices of -2.4 and -2.9.
235: The variations of the calculated fluxes were found to be less than 18\%.
236: Another study concerned the background estimation. The background is estimated using the method
237: of direct integration, calculating a background map using events over a 2hr time
238: interval~\citep{2003ApJ...595..803A}. To account for possible contamination
239: of the background from signal events
240: two prominent regions in the Milagro sky map are excluded when doing this:
241: a 2 by 2 degree region around the Crab Nebula location and a $\pm$2.5 degree region
242: around the Galactic plane. To investigate possible systematic effects in the
243: background estimation the size of the region of exclusion around the Galactic plane was increased
244: to $\pm$5 degrees. The flux variations were found to be less than 7\%.
245: These systematic errors are added in quadrature to a
246: 30\% systematic error which is derived from the underestimation of the Milagro
247: trigger rate by cosmic ray shower simulations~\citep{2007ApJ...658L..33A}.
248:
249:
250:
251: \section{Results}
252:
253: \clearpage
254: \begin{deluxetable}{ccccc}
255: \tablewidth{5.5in}
256: \tabletypesize{\small} % can be \small \tiny \scriptsize and \footnotesize
257: \tablecaption{Gamma-ray emission from the Galactic plane around 15 TeV.\label{tab:fluxes}}
258:
259: \tablehead{
260: \colhead{\bf Region} & \colhead{\bf Statistical } & \multicolumn{3}{c}{\bf Diffuse Flux ($\bf \times 10^{-13}TeV^{-1}cm^{-2}s^{-1}sr^{-1}$)} \\
261: \bf for $\bf |b|<2^0$ & \colhead{\bf Significance } & & \multicolumn{2}{c}{\bf GALPROP}\\
262: ($\bf l$, \bf deg) & {$\bf\sigma$ } & {\bf Milagro\tablenotemark{a}} & \bf optimized & \bf conventional
263: % & & & & \\
264: }
265: \startdata
266: $30<l<65$ & 5.1 & $23.1\pm 4.5^{+7.0}_{-8.0}$ & 20.0 & 4.9\\
267: $65<l<85$ & 8.6 & $21.8\pm 2.5^{+7.2}_{-7.8}$ & 10.2 & 2.7\\
268: $85<l<110$ & 1.3 & $<7.1$ (95\% c.l.) & 5.8 & 1.3\\
269: $136<l<216$ & 0.8 & $<5.7$ (95\% c.l.) & 3.1 & 0.9\\
270: \enddata
271: \tablenotetext{a}{The first error represents the statistical, the second the systematic uncertainty. See text for details.}
272: \end{deluxetable}
273: \clearpage
274:
275: The upper plot in Figure~\ref{fig:lon_prof} shows the Galactic longitude
276: profile of the $\gamma$-ray emission as measured by Milagro before and after source contributions
277: are subtracted.
278: A $\gamma$-ray flux increase towards the Galactic center is visible,
279: as well as the Cygnus region ($l\in[65^\circ,85^\circ]$) with a ``bump''
280: in the flux profile, even after source contributions are subtracted.
281: The source-subtracted data points in the lower plot of Figure~\ref{fig:lon_prof} are overlaid with the $\gamma$-ray
282: emission profiles as predicted by the optimized GALPROP
283: model, version 53\_6102129RG~\citep{2004A&A...422L..47S,2004ApJ...613..956S,2008arXiv0804.1774P}.
284: The most significant
285: discrepancy between model predictions and data appears in the Cygnus region.
286: Table~\ref{tab:fluxes} lists the emission in four different regions with the
287: Cygnus region being the most significant, 8.6~$\sigma$ above the background.
288: The predictions of the optimized and conventional GALPROP model are also given in Table~\ref{tab:fluxes}.
289: The measured diffuse flux from the Cygnus region is two (eight) times higher
290: than the optimized (conventional) GALPROP prediction. The measured flux from the
291: inner Galaxy ($l\in[30^\circ,65^\circ]$) is consistent with the optimized model and about five times higher than
292: predicted by the conventional model.
293: Calculating the ratio of the flux measured in the inner Galaxy to the flux
294: measured in the Cygnus region cancels
295: systematic detector effects like the underestimation of the trigger rate.
296: The flux ratio in the data is calculated to be 1.1$\pm$0.2 (stat.) compared
297: to a ratio of 2.0 predicted by GALPROP.
298: For the regions $l\in[85^\circ,110^\circ]$ and $l\in[136^\circ,216^\circ]$, flux upper limits
299: are quoted since the statistical significances above the background are less than two
300: standard deviations.
301:
302: The energy spectra as predicted by the optimized GALPROP model in the Cygnus region
303: and in the inner Galaxy, $l\in[30^\circ,65^\circ]$, are shown in Figure~\ref{fig:spectra}
304: together with the EGRET and Milagro measurements. As can be seen, at Milagro energies the
305: dominating GALPROP contribution to the diffuse flux is due to inverse Compton scattering of
306: cosmic-ray electrons off the cosmic microwave background (CMB).
307: The electron injection spectrum of the optimized GALPROP model is chosen such that the
308: diffuse $\gamma$-ray emission spectrum matches the EGRET measurement at GeV energies.
309: The injection spectrum is a broken power law with a spectral index of -1.5 below 20 GeV
310: and a spectral index of -2.42 above 20 GeV extending to a maximum electron energy of 1000 TeV
311: with an electron flux of 1.4$\times 10^{-9}$MeV$^{-1}$cm$^{-2}$s$^{-1}$sr$^{-1}$ at
312: 34.5 GeV~\citep{2004A&A...422L..47S,2004ApJ...613..956S}.
313:
314: Figure~\ref{fig:lon_prof30_65_85_110} shows
315: the Galactic latitude profiles for $b\in[-10^\circ,10^\circ]$
316: in three regions between Galactic longitude $l=30$ and $l=110$.
317: Gaussian fits (not shown) to the data distributions
318: for $l\in[30^\circ,65^\circ]$ (left plot) and $l\in[65^\circ,85^\circ]$ (middle plot) yield values for the mean consistent with
319: $b=0^\circ$ and for $\sigma$ of $0.9^\circ\pm 0.3^\circ$ and $2.0^\circ\pm0.2^\circ$, respectively.
320: Assuming a larger exclusion region around the Galactic plane in the background estimation
321: (see previous section) gives the same narrow width.
322: The emission profiles as predicted by the optimized GALPROP model are overlaid.
323: The blue line shows the total flux prediction, the green the inverse Compton, and the red the
324: pion contribution. For both the GALPROP prediction as well as the data the flux numbers become smaller and the
325: distributions become wider farther away from the Galactic center.
326: In order to compare the predicted profiles with the measured profiles, the $\chi^2$ in both regions is calculated.
327: In the inner Galaxy region ($l\in[30^\circ,65^\circ]$ and $b\in[-10^\circ,10^\circ]$) a $\chi^2$ of
328: 18.3 with 20 degrees of freedom is derived, corresponding to a probability of 57\% that the
329: $\chi^2$ for a correct model exceeds the observed one by chance.
330: In the Cygnus region, this chance probability is calculated to be $10^{-4}$.
331: The discrepancy between the model prediction of the latitude profile and the data in the Cygnus region
332: is investigated further by fitting the measured profile between $b=10^\circ$ and $b=-10^\circ$ to a linear combination of the
333: predicted pion and IC profiles, $C_{IC}\cdot p_{IC}(b)+C_{\pi}\cdot p_{\pi}(b)$. The factors $C_{IC}$ and
334: $C_{\pi}$ are varied independently between 0.1 and 10 in steps of 0.1. The minimum value
335: of $\chi^2$ is obtained for $C_{\pi}$=6.9, i.e. an increase of the pion contribution with respect
336: to the GALPROP prediction by a factor 6.9, and $C_{IC}$=0.1. The resulting chance probability is 3\%.
337: Performing the same fit to the inner Galaxy latitude profile yields an increase of the pion contribution
338: of $C_{\pi}$=5.8 and a decrease of the IC contribution of $C_{IC}$=0.1 with respect to the GALPROP prediction. The
339: chance probability of this result is 93\%, showing that there is no significant improvement between the two cases.
340:
341: Figure~\ref{fig:lon_prof30_65_85_110} also shows the Galactic latitude profile for $b\in[-10^\circ,10^\circ]$
342: in the region above Cygnus ($l\in[85^\circ,110^\circ]$, right plot). No significant enhancement near the Galactic plane is
343: visible.
344:
345:
346:
347: \section{Discussion}
348:
349: Measurements of the diffuse TeV $\gamma$-ray flux from the
350: Galactic Plane as well as its spatial distribution, the latitude and
351: longitude profiles, have been presented. The diffuse $\gamma$-ray flux
352: was compared to predictions based on both the conventional and the optimized GALPROP models.
353: In the $l\in[30^\circ,65^\circ]$ range the optimized GALPROP prediction is consistent
354: with the Milagro measurement.
355: According to the GALPROP model the diffuse emission near 15 TeV is
356: dominated by the inverse Compton component, which in turn is dominated
357: by $\sim$100 TeV electrons scattering off the CMB (60 to 70\% of the IC component,
358: see Figure~\ref{fig:spectra}).
359: Therefore, Milagro measurements can be interpreted as
360: showing the first indication of the cosmic-ray spectrum up and beyond 100 TeV
361: using the predictions of the GALPROP optimized model.
362: The propagated average electron
363: spectrum calculated by the optimized GALPROP model predicts a flux of about four times the locally
364: measured flux and extends above 100 TeV with a differential spectral index of -3.
365: An alternative explanation would be a harder nucleon injection spectrum in
366: the inner Galaxy than measured locally, but this would have to be checked against
367: local antiproton measurements~\citep{1998A&A...338L..75M,strongsref}.
368: Studies of the lateral $\gamma$-ray emission profile in the inner Galaxy show that the profile is rather narrow
369: and suggest a better agreement with the pion assumption. Better sensitivity is needed however to
370: significantly differentiate between the IC and the pion hypotheses.
371:
372:
373: The Cygnus region is the region with the most significant flux excess
374: in the Milagro spatial maps. The emission from the Cygnus region of the
375: Galaxy at longitudes $l\in[65^\circ,85^\circ]$ shows an excess by a factor
376: of two when compared to the GALPROP optimized model predictions.
377: Also for the Cygnus region, the model predictions are dominated by IC scattering of
378: electrons off the CMB (see Figure~\ref{fig:spectra}). However,
379: the measured and predicted profiles are inconsistent (see Figure~\ref{fig:lon_prof30_65_85_110}).
380: Decreasing the IC contribution (which has a broad profile distribution) and increasing the
381: pion contribution (which has a narrow profile distribution) with respect to
382: the model prediction improves the agreement in the shape,
383: reflecting that the latitude distribution measured by Milagro is rather
384: narrow compared to the distribution predicted by the optimized GALPROP model.
385:
386: The Cygnus region hosts intense star formation activity, and is abundant
387: with molecular clouds and candidate cosmic ray sources.
388: The HEGRA
389: source TeV J2032+413 is located in the area with the largest
390: concentration of molecular and atomic hydrogen in the Cygnus
391: region and spatially coincident with the Milagro source
392: MGRO J2031+41~\citep{2007ApJ...658L..33A}.
393: For TeV J2032+413 an association has been proposed with
394: Cygnus OB2, the largest cluster of more than 2700 identified
395: young stars at 5000 light year distance~\citep{2005A&A...431..197A,2007PhRvD..76f7301A}.
396: Very close to TeV J2032+4130, the VLA has detected a weak non-thermal shell supernova
397: remnant~\citep{2006astro.ph.11731B} and more
398: recent XMM-Newton observations have revealed the presence of an extended
399: X-ray source co-located with TeV J2032+4130, which could be produced by an
400: unknown population of faint X-ray sources~\citep{2007A&A...469L..17H}.
401: In order to
402: explain the X-ray and TeV emissions from TeV J2032+413, \citet{2007A&A...469L..17H}
403: proposed both a hadronic and a leptonic scenario. In the hadronic
404: scenario, the X-ray emission would be produced by the synchrotron emission
405: of secondary electrons and the emission above 10 TeV should show a hard
406: spectrum. If the X-ray and TeV emissions were instead of leptonic origin,
407: the spectrum above 10 TeV should be rather soft due to the unavoidable
408: Klein-Nishina suppression of the inverse Compton cross section.
409:
410: If cosmic rays are injected into the Galactic interstellar
411: medium by any of these candidate counterparts to TeV J2032+4130 and
412: MGRO J2031+41, then the newly accelerated cosmic rays interact with
413: the local gas and produce $\gamma$ rays within 100 pc from the source
414: with spectra which
415: might be significantly different from the average $\gamma$-ray spectrum,
416: because the cosmic-ray spectrum has not yet been steepened by
417: diffusion~\citep{1996A&A...309..917A}. Assuming a distance of about
418: 1 kpc, the extension of the entire Cygnus region is about 300 pc, and thus
419: a single accelerator might influence strongly the entire region.
420: Assuming the diffusion process to be energy dependent, the emission from a
421: molecular cloud close to such an accelerator could be due to high energy protons
422: which have been accelerated first, have left the remnant first,
423: and have already reached the cloud, as they diffuse faster.
424: Thus this emission would have a flatter spectrum and could produce
425: VHE $\gamma$ rays in the Milagro energy range during
426: the first 10,000 years after the SN
427: explosion~\citep{2007ApJ...665L.131G,2007arXiv0705.3854M}.
428: Leptonic
429: mechanisms for the production of VHE $\gamma$ rays are disfavored by
430: the fast cooling of such highly energetic electrons.
431: Assuming a cosmic-ray source of total energy ${10}^{51}$ ergs having exploded less than
432: 10,000 years ago in the ISM and a molecular cloud of total mass greater than
433: ${10}^4$ and less than ${10}^5$ solar masses~\citep{1994ApJS...95..419D},
434: the hard spectrum cosmic ray nucleons scattering off the targets in the Cygnus region would
435: produce a flux between ${10}^{-15}$ and ${10}^{-14}$ TeV$^{-1}$cm$^{-2}$s$^{-1}$
436: near 15 TeV
437: (see Figure 1 of~\citet{2007ApJ...665L.131G}).
438: Calculated from Table~\ref{tab:fluxes}, the emission measured by Milagro near 15 TeV not explained
439: by GALPROP is roughly $ 2.9 \times{10}^{-14}$ TeV$^{-1}$cm$^{-2}$s$^{-1}$, meaning that
440: only a few strong young accelerators in the Cygnus region are needed to explain the
441: excess emission measured by Milagro.
442:
443: The results herein and previously presented by the Milagro Collaboration
444: provide further evidence for the existence of cosmic-ray accelerators in the Cygnus
445: region of the Galaxy, favoring hadronic production mechanisms~\citep{2007PhRvD..75h3001B,2007PhRvD..76f7301A,2006astro.ph.11731B}.
446: If the emission originates from hadronic interactions of hard spectrum cosmic rays, a
447: corresponding flux of neutrinos will arise
448: from decay of charged pions, and neutrino detectors such as Icecube could provide the conclusive
449: probe of proton acceleration in the Galaxy~\citep{2007PhRvD..75h3001B,2007ApJ...665L.131G}.
450:
451: Experiments like GLAST (with its improved sensitivity and angular resolution with respect to
452: EGRET) will be able to separate the truly
453: diffuse $\gamma$-ray emission from a potentially unresolved source component, and probe the spatial distribution
454: of the diffuse emission at GeV energies (the ``GeV excess'').
455: Experiments like the proposed High Altitude Water Cherenkov (HAWC)
456: detector will be able to constantly survey large regions of the sky,
457: in particular the Galactic plane, at $\gamma$-ray energies up to $\sim$100 TeV with 10 to 15 times the sensitivity of Milagro.
458: Because of its more southerly location it will also be more sensitive to Galactic plane regions below $l=65^\circ$.
459: This will put stricter constraints on models like GALPROP and thus provide crucial information
460: about the propagation of cosmic rays above 100 TeV.
461:
462:
463: \acknowledgements
464:
465: We thank Scott Delay, Michael Schneider, and Owen Marshall for their dedicated efforts
466: on the Milagro experiment. We also gratefully acknowledge the financial support of
467: the National Science Foundation (under grants
468: PHY-0245234, %UCI; previous is PHY-0070933
469: -0302000, %UMD
470: -0400424, %Milagro Operations; previous is PHY-0075326
471: -0504201, %NYU current; previous are PHY-9901496 & PHY-0206656
472: -0601080, %UCSC; previous are PHY-0070927 & PHY-0245143
473: and
474: ATM-0002744), %UNH
475: the Department of Energy (Office of High Energy Physics), Los
476: Alamos National Laboratory, the University of California,
477: and the Institute for Geophysics and Planetary
478: Physics at Los Alamos National Laboratory.
479: I.~V.~Moskalenko acknowledges partial support from the NASA APRA grant.
480:
481:
482: \begin{thebibliography}{3}
483:
484: % Milagro Cygnus Paper
485: \bibitem[Abdo et al.(2007a)]{2007ApJ...658L..33A} Abdo, A.~A., et al.\ 2007a, \apjl, 658, L33
486: % Milagro Galactic Survey Paper
487: \bibitem[Abdo et al.(2007b)]{2007ApJ...664L..91A} Abdo, A.~A., et al.\ 2007b, \apjl, 664, L91
488: % GEANT reference
489: \bibitem[Agostinelli et al. (2003)]{geant} Agostinelli, S., et al. 2003, NIM A 506 2003, 250
490: % Aharonian%Atoyan 1996, young acclerators vs sea
491: \bibitem[Aharonian \& Atoyan(1996)]{1996A&A...309..917A} Aharonian, F.~A.,
492: \& Atoyan, A.~M.\ 1996, \aap, 309, 917
493: % Aharonian%Atoyan 2000
494: \bibitem[Aharonian \& Atoyan(2000)]{2000A&A...362..937A} Aharonian, F.~A.,
495: \& Atoyan, A.~M.\ 2000, \aap, 362, 937
496: % The unidentified TeV source (TeV J2032+4130) and surrounding field: Final HEGRA IACT-System results
497: \bibitem[Aharonian et al.(2005a)]{2005A&A...431..197A} Aharonian, F., et
498: al.\ 2005a, \aap, 431, 197
499: % More HESS
500: \bibitem[Aharonian et al.(2005b)]{2005Sci...307.1938A} Aharonian, F., et
501: al.\ 2005b, Science, 307, 1938
502: % SN reference
503: \bibitem[Aharonian et al.(2006a)]{2006A&A...449..223A} Aharonian, F., et
504: al.\ 2006a, \aap, 449, 223
505: % HESS Nature Galactic Center Ridge
506: \bibitem[Aharonian et al.(2006b)]{2006Natur.439..695A} Aharonian, F., et
507: al.\ 2006b, \nat, 439, 695
508: % Even more HESS
509: \bibitem[Aharonian et al.(2006c)]{2006ApJ...636..777A} Aharonian, F., et
510: al.\ 2006c, \apj, 636, 777
511: % Tibet anisotropy
512: \bibitem[Amenomori et al.(2006)]{2006Sci...314..439A} Amenomori, M., et
513: al.\ 2006, Science, 314, 439
514: % Icecube: Neutrino flux from cosmic ray accelerators in the Cygnus spiral arm of the Galaxy
515: \bibitem[Anchordoqui et al.(2007)]{2007PhRvD..76f7301A} Anchordoqui, L.,
516: Halzen, F., Montaruli, T., \& O'Murchadha, A.\ 2007, \prd, 76, 067301
517: % Background rejection technique reference
518: \bibitem[Atkins et al.(2003)]{2003ApJ...595..803A} Atkins, R., et al.\
519: 2003, \apj, 595, 803
520: % Milagro detector reference
521: \bibitem[Atkins et al.(2004)]{2004ApJ...608..680A} Atkins, R., et al.\
522: 2004, \apj, 608, 680
523: % Milagro evidence for TeV gamma ray emission from the galactic plane
524: \bibitem[Atkins et al.(2005)]{2005PhRvL..95y1103A} Atkins, R., et al.\
525: 2005, Physical Review Letters, 95, 251103
526: % underestimation of EGRET sensitivity, opposite conclusion of Stecker et al.
527: \bibitem[Baughman et al.(2007)]{2007arXiv0706.0503B} Baughman, B.~M.,
528: Atwood, W.~B., Johnson, R.~P., Porter, T.~A., \& Ziegler, M.\ 2007, ArXiv
529: e-prints, 706, arXiv:0706.0503
530: % Disecting Cygnus
531: \bibitem[Beacom \& Kistler(2007)]{2007PhRvD..75h3001B} Beacom, J.~F., \&
532: Kistler, M.~D.\ 2007, \prd, 75, 083001
533: % Bertsch
534: \bibitem[Bertsch et al.(1993)]{1993ApJ...416..587B} Bertsch, D.~L., Dame,
535: T.~M., Fichtel, C.~E., Hunter, S.~D., Sreekumar, P., Stacy, J.~G., \&
536: Thaddeus, P.\ 1993, \apj, 416, 587
537: % TeV J2032+4130: a not-so-dark Accelerator? (VLA)
538: \bibitem[Butt et al.(2006)]{2006astro.ph.11731B} Butt, Y.~M., Combi, J.~A.,
539: Drake, J., Finley, J.~P., Konopelko, A., Lister, M., Rodriguez, J., \&
540: Shepherd, D.\ 2006, ArXiv Astrophysics e-prints, arXiv:astro-ph/0611731
541: % Sabrina and Brenda
542: \bibitem[Casanova \& Dingus(2007)]{2007arXiv0711.2753C} Casanova, S., \&
543: Dingus, B.~L.\ 2007, Astroparticle Physics (in press), ArXiv e-prints, 711, arXiv:0711.2753
544: % dark matter de Boer
545: \bibitem[de Boer et al.(2005)]{2005A&A...444...51D} de Boer, W., Sander,
546: C., Zhukov, V., Gladyshev, A.~V., \& Kazakov, D.~I.\ 2005, \aap, 444, 51
547: % for Sabrina's calculation: cloud masses
548: \bibitem[Dobashi et al.(1994)]{1994ApJS...95..419D} Dobashi, K., Bernard,
549: J.-P., Yonekura, Y., \& Fukui, Y.\ 1994, \apjs, 95, 419
550: % Gabici and Aharonian
551: \bibitem[Gabici \& Aharonian(2007)]{2007ApJ...665L.131G} Gabici, S., \&
552: Aharonian, F.~A.\ 2007, \apjl, 665, L131
553: % Gralewicz
554: \bibitem[Gralewicz et al.(1997)]{1997A&A...318..925G} Gralewicz, P.,
555: Wdowczyk, J., Wolfendale, A.~W., \& Zhang, L.\ 1997, \aap, 318, 925
556: % Corsika
557: \bibitem[Heck et al. (1998)]{corsika} Heck, D. et al., Report FZKA 6019, 1998
558: % XMM-Newton observations
559: \bibitem[Horns et al.(2007)]{2007A&A...469L..17H} Horns, D., Hoffmann,
560: A.~I.~D., Santangelo, A., Aharonian, F.~A., \& Rowell, G.~P.\ 2007, \aap,
561: 469, L17
562: % the Hunter bible
563: \bibitem[Hunter et al.(1997)]{1997ApJ...481..205H} Hunter, S.~D., et al.\
564: 1997, \apj, 481, 205
565: % Observations of the Unidentified TeV gamma-Ray Source TeV J2032+4130 with the Whipple Observatory 10 m Telescope
566: \bibitem[Konopelko et al.(2007)]{2007ApJ...658.1062K} Konopelko, A., et
567: al.\ 2007, \apj, 658, 1062
568: % Diffuse galactic gamma rays, cosmic-ray nucleons and antiprotons
569: \bibitem[Moskalenko et al.(1998)]{1998A&A...338L..75M} Moskalenko, I.~V., Strong, A.~W., \& Reimer, O.\ 1998, \aap, 338, L75
570: % Understanding limitations in the determination of diffuse Galactic gamma-ray emission
571: \bibitem[Moskalenko et al.(2007a)]{2007NuPhS.173...44M} Moskalenko, I.~V., Digel, S.~W., Porter, T.~A.,
572: Reimer, O., \& Strong, A.~W.\ 2007a, Nuclear Physics B Proceedings Supplements, 173, 44
573: % Hadronic Gamma Rays from SNR
574: \bibitem[Moskalenko et al.(2007b)]{2007arXiv0705.3854M} Moskalenko, I.~V., Porter, T.~A., Malkov, M.~A.,
575: \& Diamond, P.~H.\ 2007b, arXiv e-prints, 705, arXiv:0705.3854
576: % Porter&Protheroe
577: \bibitem[Porter et al.(2008)]{2008arXiv0804.1774P} Porter, T.~A.,
578: Moskalenko, I.~V., Strong, A.~W., Orlando, E.,
579: %\& Bouchet, L.\ 2008, ArXiv e-prints, 804, arXiv:0804.1774.
580: \& Bouchet, L.\ 2008, \apj 682, 400
581: \bibitem[Porter \& Protheroe(1997)]{1997JPhG...23.1765P} Porter, T.~A., \&
582: Protheroe, R.~J.\ 1997, Journal of Physics G Nuclear Physics, 23, 1765
583: % Prodanovic
584: \bibitem[Prodanovi{\'c} et al.(2007)]{2007APh....27...10P} Prodanovi{\'c},
585: T., Fields, B.~D., \& Beacom, J.~F.\ 2007, Astroparticle Physics, 27, 10
586: % Stecker, EGRET systematics
587: \bibitem[Stecker et al.(2008)]{2008APh....29...25S} Stecker, F.~W., Hunter,
588: S.~D., \& Kniffen, D.~A.\ 2008, Astroparticle Physics, 29, 25
589: % Comptel points
590: \bibitem[Strong et al.(1999)]{1999ApL&C..39..209S} Strong, A.~W., Bloemen,
591: H., Diehl, R., Hermsen, W., \& Sch{\"o}nfelder, V.\ 1999, Astrophysical
592: Letters Communications, 39, 209
593: % Strong and Moskalenko
594: % GALPROP general citation
595: \bibitem[Strong et al.(2000)]{2000ApJ...537..763S} Strong, A.~W.,
596: Moskalenko, I.~V., \& Reimer, O.\ 2000, \apj, 537, 763
597: % GALPROP tuned to EGRET
598: \bibitem[Strong et al.(2004a)]{2004A&A...422L..47S} Strong, A.~W.,
599: Moskalenko, I.~V., Reimer, O., Digel, S., \& Diehl, R.\ 2004a, \aap, 422, L47
600: % GALPROP tuned to EGRET
601: \bibitem[Strong et al.(2004b)]{2004ApJ...613..956S} Strong, A.~W.,
602: Moskalenko, I.~V., \& Reimer, O.\ 2004b, \apj, 613, 956
603: \bibitem[Strong et al.(2007)]{strongsref} Strong, A.~W., Moskalenko, I.~V.,
604: \& Ptuskin, V.~S.\ 2007, Annu.~Rev.~Nucl.~Part.~Sci., 57, 285
605:
606: \end{thebibliography}
607:
608: \clearpage
609:
610: \begin{figure*}[width=4.5in,p]
611: \plotone{f1a.eps}
612: \plotone{f1b.eps}
613: \caption{Galactic longitude profile of the $\gamma$-ray emission around 15 TeV in
614: the Galactic plane as measured by Milagro. Upper plot: Red data points with dashed error bars
615: -- no subtraction of source contributions,
616: black data points -- after subtraction of source contributions.
617: Lower plot: Source-subtracted profile overlaid with prediction of the optimized GALPROP model --
618: the red line is the pion contribution, the green line the IC contribution, and the blue line
619: represents the total flux prediction between Galactic latitudes $\pm$ 2 degrees.
620: There are no data points in
621: the region of longitude $l\in[-144^\circ,29^\circ]$ because it is below the Milagro horizon. The
622: region $l\in[111^\circ,135^\circ]$ is excluded because the analysis method is insensitive here (see
623: text for details).}
624: \label{fig:lon_prof}
625: \end{figure*}
626:
627:
628: \clearpage
629:
630: \begin{figure*}[width=6in,p]
631: \plottwo{f2a.eps}{f2b.eps}
632: \caption{Gamma-ray spectra of the diffuse emission as predicted by the optimized GALPROP model for the
633: Galactic plane -- left plot: inner Galaxy ($l\in[30^\circ,65^\circ]$), right plot:
634: Cygnus region ($l\in[65^\circ,85^\circ]$). The red
635: bars represent EGRET data, the black bar the Milagro measurement, where the length of the bar represents
636: the statistical uncertainty only. The dark blue line represents the total diffuse flux predicted by the
637: optimized GALPROP model, the dark gray line the extragalactic background, and the light blue line the
638: bremsstrahlung component. The two contributions at Milagro energies are shown as red line, the pion
639: contribution, and green line, the total IC contribution. The green dashed line shows the dominant IC contribution
640: from scattering of electrons off the cosmic microwave background, which amounts to about 60 to 70\% of the IC
641: component at Milagro energies. Other IC contributions which are less important, such as infrared and optical,
642: are not shown separately.}
643: \label{fig:spectra}
644: \end{figure*}
645:
646: \clearpage
647:
648: \begin{figure*}[width=4.5in,p]
649: \plotone{f3.eps}
650: \caption{Source-subtracted Galactic latitude profile of the $\gamma$-ray emission around 15 TeV in
651: the inner Galaxy (left plot), in the Cygnus region (middle plot), and in the region above Cygnus (right plot) as measured by Milagro (points
652: with errors) and predicted by the optimized GALPROP model. The blue curve is the total $\gamma$-ray flux,
653: the red curve the pion and the green curve the IC contribution.}
654: \label{fig:lon_prof30_65_85_110}
655: \end{figure*}
656:
657: \clearpage
658:
659: \end{document}
660:
661:
662: