1:
2: \documentstyle[12pt,epsfig,amssymb]{article}
3:
4: %\pagestyle{empty}
5: \renewcommand
6: \baselinestretch{2}
7:
8: %
9: %%%%%%%%%%%%%%%%%%
10: % STANDARD LATEX %
11: % MACROS: EPSFIG %
12: % AMSSYMB %
13: %%%%%%%%%%%%%%%%%%%
14: %
15: %\textheighft 23.cm
16: \textheight 22.cm
17: \textwidth 15cm
18: \oddsidemargin 0.0cm \evensidemargin 0.0cm
19: %\topmargin -0.5cm
20: \topmargin -1.5cm
21: \renewcommand{\textfraction}{0}
22: \renewcommand{\topfraction}{1}
23: %
24: %TITLEPAGE
25: %
26: \begin{document}
27: %
28: \setlength{\parskip}{0.45cm}
29: \setlength{\baselineskip}{0.75cm}
30: %
31: \begin{titlepage}
32:
33: \vspace{1cm}
34: %\begin{flushright}
35: %CERN-TH/99-01 \\
36: %{\tt astro-ph/9901004}\\
37: %\end{flushright}
38: %
39: \vspace{0.5cm}
40: \begin{center}
41: \Large
42:
43: {\bf
44: Is the diffuse Gamma Background Radiation generated by
45: Galactic Cosmic Rays?}\\
46: \vspace{0.5cm}
47: \large
48: Arnon Dar$^{a,b}$ and A. De R\'ujula$^a$
49: \\
50:
51: \vspace*{0.5cm}
52: \normalsize
53: $^a$ Theory Division, CERN, CH-1211 Geneva 23, Switzerland \\
54: $^b$ Technion, Israel Institute of Technology,
55: Haifa 32000, Israel\\
56: \vspace{0.7cm}
57: %
58: %%%%%%%%%%%%%%%%%
59: \large
60: {\bf Abstract} \\
61: \end{center}
62: \vspace*{-0.5cm}
63: \noindent
64: We explore the possibility that the diffuse gamma-ray background radiation
65: (GBR) at high galactic latitudes could be dominated by inverse Compton
66: scattering of cosmic ray (CR) electrons on the cosmic microwave background
67: radiation and on starlight from our own galaxy.
68: Assuming that the
69: mechanisms accelerating galactic CR hadrons and electrons are the same, we
70: derive simple and successful relations between the spectral indices of the
71: GBR above a few MeV, and of the CR electrons and CR nuclei above a few
72: GeV. We reproduce the observed intensity and angular dependence of the
73: GBR, in directions away from the galactic disk and centre, without
74: recourse to hypothetical extragalactic sources.
75:
76: \vspace{1.5cm}
77:
78: PACS numbers: 98.70.Sa, 98.70.Rz, 98.70.Vc.
79:
80: \vspace{1.5cm}
81:
82:
83: \end{titlepage}
84: \newpage
85: \renewcommand
86: \baselinestretch{1}
87: \normalsize
88: \section{Introduction}
89:
90:
91: The existence
92: of an isotropic, diffuse gamma background radiation (GBR)
93: was first suggested by
94: data from the SAS 2 satellite (Thompson \& Fichtel 1982).
95: The EGRET instrument on the Compton Gamma Ray Observatory
96: confirmed this finding: by removal of point sources and of the
97: galactic-disk and galactic-centre emission, and after an extrapolation
98: to zero local column density,
99: a uniformly distributed GBR was found, of alleged
100: extragalactic origin (Sreekumar et al.~1998). Above an energy of
101: $\sim\!10$ MeV, this radiation --to which we shall refer throughout simply
102: as ``the GBR''-- has a featureless spectrum, shown in Fig.~1,
103: which is very well described by a simple power-law form,
104: ${\rm dF/dE\propto E^{-\beta}}$, with $\beta\approx 2.10\pm0.03$ (Sreekumar et al.
105: 1998).
106:
107: The origin of the GBR is still unknown.
108: The published candidate sources range from the
109: quite conventional to the decisively speculative.
110: Perhaps the most conservative hypothesis for the origin of an isotropic GBR
111: is that it is extragalactic, and originates from active galaxies (Bignami et
112: al.~1979; Kazanas \& Protheroe 1983; Stecker \& Salamon 1996).
113: The fact that blazars have a $\gamma$-ray spectrum with an average index
114: $2.15\pm 0.04$, compatible with that of the GBR, supports
115: this hypothesis (Chiang \& Mukerjee 1998).
116: The possibility has also been discussed
117: that Geminga-type pulsars, expelled into
118: the galactic halo by asymmetric supernova explosions,
119: be abundant enough to explain the GBR (Dixon et al. 1998;
120: Hartmann 1995). More exotic hypotheses include a baryon-symmetric
121: universe (Stecker et al.~1971), now excluded (Cohen et al.~1998), primordial black hole evaporation (Page \& Hawking 1976; Hawking 1977),
122: supermassive black holes
123: formed at very high redshift (Gnedin \& Ostriker 1992), annihilation of
124: weakly interactive big-bang remnants (Silk \& Srednicki 1984; Rudaz
125: \& Stecker 1991), and a long etc.
126:
127: However, the EGRET GBR data in directions above the galactic disk and centre
128: show a significant deviation from isotropy, correlated with the
129: structure of our galaxy and our position relative to its centre
130: (Dar et al.~1999). This advocates a local (as
131: opposed to cosmological) origin for the GBR.
132: Indications of a large galactic
133: contribution to the GBR at large latitudes were independently
134: found by Dixon et al. (1998) by means of a wavelet-based ``non-parametric''
135: approach that makes no reference to a particular model.
136: Strong \& Moskalenko (1998) and Moskalenko \& Strong (2000)
137: also found that the contribution of inverse Compton scattering of galactic
138: cosmic ray electrons to the diffuse $\gamma$-ray
139: background is presumably much larger than previously thought.
140: In this paper we go one step further and explore in detail
141: the possibility (Dar et al.~1999)
142: that the diffuse gamma-ray background radiation
143: at high galactic latitudes could be dominated by inverse Compton
144: scattering of cosmic ray (CR) electrons on the cosmic microwave background
145: radiation and on starlight from our own galaxy.
146: In Section 2 we briefly review
147: the GBR data and the evidence for its correlation with our
148: position in the Galaxy.
149:
150:
151: The CR-proton and CR-electron spectra
152: are briefly reviewed in Section 3.
153: The origin, spectrum and composition
154: of non-solar cosmic ray protons and nuclei have
155: been debated for almost a century.
156: The measurements now extend over
157: some 30 orders of magnitude in flux and some 15
158: orders of magnitude in energy, up to an astonishing
159: ${\rm E}\!\sim\! 3\!\times\!10^{11}$ GeV
160: (Bird et al.~1995, Takeda et al.~1998, Berezinskii et al.~1990
161: and references therein). Above
162: $\sim\! 5$ GeV,
163: this spectrum has also a power-law form ${\rm E^{-\beta}}$,
164: with two small
165: variations in the ``index'' $\beta$ at the so-called
166: ``CR knee'' and ``CR ankle''. The local spectrum
167: of CR electrons, shown in Fig.~2,
168: is much harder to measure; it is only known up to
169: $\sim\! 10^3$ GeV and, above $\sim\! 5$ GeV, it is also well described
170: by a simple power law.
171:
172: In Sections 4 and 5 we discuss relations between the indices
173: of the GBR and the CR electron and proton spectra. In so doing,
174: we make few and very simple
175: assumptions: that the mechanism accelerating CR hadrons and
176: CR electrons is the same (a moving magnetic ``mirror''),
177: that the locally-measured
178: electron spectrum is representative of its average form
179: throughout the Galaxy, that above a certain energy, inevitably,
180: the electron spectrum is modulated by inverse
181: Compton scattering on starlight and on the
182: microwave background radiation, and that the GBR
183: is dominated by the resulting Compton up-scattered
184: photons. This allows one to derive, successfully, the GBR index from
185: the electron index and the electron index from the
186: proton index. The GBR index,
187: as observed by EGRET, is uncannily
188: directionally uniform. We interpret this fact as strong support
189: for our simple assumptions.
190:
191: In Section 6 we tackle a more difficult and potentially
192: controversial subject: the origin and magnitude of the
193: GBR. In a sense, our proposed explanation
194: --that the GBR originates from inverse Compton scattering
195: in our own
196: galaxy (Dar et al.~1999) -- is more conservative than any of the
197: previously suggested origins.
198:
199: The non-conventional aspect of our hypothesis is that,
200: in order to reproduce the observed intensity of the GBR,
201: we must assume the scale height of our galaxy's
202: CR-electron distribution to be almost twice the traditionally-accepted
203: upper limit. Because of this, in Section 6, we briefly
204: review the basis of the conventional wisdom and our
205: critical view of it, whose main points are the following.
206: Moskalenko, Strong
207: and their collaborators have developed
208: a very detailed understanding of the CR, radio and $\gamma$
209: observations of our galaxy.
210: To fit the data, their models require a freely parametrized
211: reacceleration of electrons, presumably by the motion
212: of turbulent magnetic fields (e.g., Seo \& Ptuskin, 1994).
213: Strong \& Moskalenko (1998) introduce a cutoff ${\rm z_h}$
214: for the height above the galactic plane above which
215: cosmic rays freely escape. They find an
216: upper limit ${\rm z_h\!<\!12}$ kpc,
217: on the basis of a fit to the ${\rm ^{10}Be/^9 Be}$ ratio observed by
218: Ulysses (Connell 1998). This result
219: is ``soft'': twice the upper limit would still be compatible with
220: the ensemble of data (Lukasiak et al.~1994). Moreover, the galactic
221: CR proton distribution extracted from a fit to EGRET $\gamma$-ray
222: data, actually favours (Strong \& Moskalenko 1998) an ad hoc distribution of
223: CR sources that is not as well localized in the disk as the
224: conventional supernova-remnant sources are (Webber 1997),
225: even if ${\rm z_h=20}$ kpc or more.
226: This point, and the necessity to invoke CR reacceleration,
227: indicate that scale heights of the CR electron distribution
228: in excess of the 12 kpc ``upper limit'' may not be out of the
229: question. Our results are optimized by a
230: scale height of roughly 20 kpc.
231: Such a large scale height is not in contradiction
232: with radio synchrotron-emission from our galaxy if the galactic disk
233: and its magnetic field are embedded in a larger magnetic halo
234: with a much weaker field.
235:
236:
237: In studying the possibility that the diffuse GBR
238: is not extragalactic, one has two choices.
239: The first is to extend to high galactic latitudes
240: the elaborate models (with many parameters,
241: reacceleration, and ad hoc modifications
242: of the CR-proton and CR-electron energy and source distributions)
243: that have been developed to describe the intricate nature of the
244: observations at low galactic latitudes (Strong \& Moskalenko 1998;
245: Moskalenko \& Strong 2000).
246: The second is to adopt our very naive set of hypotheses
247: and employ a simple cosmic-ray model with, by conventional
248: standards, a large scale height for CR-electrons.
249: Models of this type (Dar \& Plaga 1999), wherein cosmic ray sources are
250: directly injected at high galactic
251: latitudes, have
252: actually been proposed\footnote{The injector agents
253: would be highly relativistic
254: jets from the birth of compact objects in supernova explosions,
255: leading to a CR population permeating a magnetized region
256: of galactic-halo proportions and constituting a putative
257: solution to the problem of the origin of the highest-energy
258: cosmic rays,
259: a qualitative description of the nuclear CR spectrum, and a
260: possible explanation of jetted gamma-ray bursts.}.
261:
262:
263: In Section 7 we discuss the magnitude and angular-dependence
264: of the two dominant contributions to the GBR within our model:
265: inverse Compton scattering of galactic CR-electrons
266: off the cosmic background radiation and starlight.
267: In Section 8 we compute the small additive effect
268: of sunlight, and in Section 9 we estimate the contribution
269: from external galaxies, which is also sub-dominant.
270: In Section 10 we compare our predictions with the data on the
271: intensity and the angular dependence of the GBR. The results
272: are very satisfactory and, within our model, lead to
273: the conclusion that the GBR can be dominated by the
274: emission from our own galaxy. We summarize our conclusions
275: and predictions in Section 11.
276:
277: \section{The GBR data}
278: We call ``the GBR'' the
279: diffuse emission observed by EGRET
280: by masking the galactic plane at latitudes
281: $\rm{|b|\le 10^o}$, as well as the galactic centre
282: at $\rm{|b|\le 30^o}$ for longitudes $\rm{|l|\le 40^o}$,
283: and by extrapolating to zero column density, to eliminate the $\pi^0$
284: and bremsstrahlung contributions to the observed radiation and
285: to tame the model-dependence of the results.
286: Outside the mask, the GBR flux integrated over all directions in the
287: observed energy range of ${\rm 30}$ MeV to ${\rm 120~GeV}$, shown in
288: Fig.~1, is well described by a power law:
289: \begin{equation}
290: {\rm {dF_\gamma\over dE}\simeq (2.74\pm 0.11)\times 10^{-3} \left
291: [E\over MeV \right]^{-2.10\pm 0.03}
292: ~cm^{-2}~s^{-1}~sr^{-1}~MeV^{-1}}\; .
293: \label{photons}
294: \end{equation}
295: The overall magnitude
296: in Eq.~(\ref{photons}) is sensitive to the model used to subtract
297: the foreground (Sreekumar et al.~1998; Strong et al.~1998), but the
298: spectral index is not.
299: The EGRET data are given in Sreekumar et al.~(1998) for 36 $\rm{(b,l)}$ domains, 9 values for each half-hemisphere.
300: The spectral index is, within errors, extremely directionally uniform,
301: as shown in Fig.~3, where we have plotted the EGRET results
302: as functions of $\theta$, the observation angle
303: relative to the direction to the galactic centre
304: (${\rm \cos\theta=\cos\,[b]\, \cos\,[l]}$).
305: The normalization
306: is less homogeneous, but in directions well above the galactic disk and
307: away from the galactic-centre region it
308: has been found to be consistent with
309: a normal distribution around the mean value:
310: thus the claim of a possible extragalactic origin (Sreekumar et al. 1998).
311:
312: In Fig.~4 we have plotted, as a function of $\theta$, the EGRET GBR counting-rate above 100 MeV. This figure clearly shows,
313: in three out of the four quarters of the celestial sphere, an
314: increase of the counting rate towards the galactic
315: centre. How significant is this effect? Let ${\rm \bar\chi^2\equiv\chi^2/d.o.f.}$
316: be the ``reduced'' $\chi^2$ per degree of freedom.
317: The $\bar\chi^2$ value for constant flux
318: is 2.6: very unsatisfactory. A best fit of the form
319: ${\rm F=F_0+F_1\,(1-\cos\theta)}$ yields
320: ${\bar\chi^2=1.3}$, a very large amelioration (for higher
321: polynomials in $\cos\theta$ the higher-order coefficients are
322: compatible with zero: the fit does not significantly improve).
323: Note also that at angles with $\cos\theta$ larger than its mean value
324: $\langle\cos\theta\rangle\! =\! 0.0246$ (${\rm \theta<88.6^o}$), 10 out of the 12
325: data points are above the average flux, while at
326: angles with ${\rm \theta>88.6^o}$, 18 out of the 24 data points
327: are below the average. The probability for a
328: uniform distribution to produce this large or larger a
329: fluctuation is $1.5\times 10^{-4}$.
330:
331:
332: Even in directions pointing to the galactic disk
333: and the galactic centre,
334: EGRET data on $\gamma$-rays above 1 GeV show an excess over
335: the expectation
336: from galactic cosmic-ray production of $\pi^0$'s (Pohl \& Esposito
337: 1998). Electron bremsstrahlung in gas is not the source
338: of the 1--30 MeV inner-Galaxy $\gamma$-rays observed by
339: COMPTEL (Strong et al.~1997), since their galactic latitude distribution is
340: broader than that of the gas. These findings also imply that inverse Compton scattering
341: may be much more important than previously believed (Strong \& Moskalenko 1998; Moskalenko and Strong, 2000;
342: Dar et al.~1999).
343:
344:
345: \section{The CR data}
346:
347: The cosmic ray nuclei have a power-law spectral flux ${\rm
348: dF/dE\propto E^{-\beta}}$ with an index $\beta$ that changes at two
349: break-point energies. In the interval ${\rm 10^{10}~eV < E <
350: E_{knee}}$ $\sim 3 \times 10^{15}$ eV, protons constitute $\sim 96\%$
351: of the CRs at fixed energy per nucleon, and their flux is
352: (Berezinskii et al.~1990, and references therein):
353: \begin{equation}
354: {\rm {dF_p\over dE}\simeq 1.8 \left
355: [E \over GeV \right]^{-2.70\pm 0.05}
356: ~cm^{-2}~s^{-1}~sr^{-1}~GeV^{-1}}.
357: \label{protons}
358: \end{equation}
359: In the interval
360: $~{\rm E_{knee} < E < E_{ankle}}$ $\sim 3 \times 10^{18}$ eV,
361: the spectrum steepens from $\beta_1\sim 2.7$
362: to $\beta_2\sim 3.0$, flattening again to $\beta_3\sim 2.5$
363: above $\rm{ E_{ankle}}$.
364:
365: The CR flux of electrons (Prince 1979; Nishimura et al.~1980; Tang 1984; Golden et al.~1984;
366: Evenson \&
367: Meyers 1984; Golden et al.~1994; Ferrando et al.~1996; Barwick et al.~1998;
368: Wiebel-Sooth \& Biermann 1998), shown in Fig.~2,
369: is well fitted, from ${\rm E\sim 10~GeV}$ to $\sim 2$ TeV by:
370: \begin{equation}
371: {\rm {dF_e\over dE}\simeq (2.5\pm 0.5)\times 10^5 \left
372: [E \over MeV \right]^{-3.2\pm 0.10}
373: ~cm^{-2}~s^{-1}~sr^{-1}~MeV^{-1}}.
374: \label{electrons}
375: \end{equation}
376: The terrestrial and solar magnetic fields and the solar wind modify
377: the electron
378: spectrum below ${\rm E\sim 10}$ GeV, so that
379: the direct observations at those energies
380: may deviate from the local interstellar spectral shape.
381:
382: Cosmic ray electrons undergo inverse Compton scattering
383: (ICS) off the ambient photon baths:
384: starlight and the cosmic background radiation.
385: The spectral indices of the GBR and
386: electron spectra
387: can be very simply and successfully related (Dar et al.~1999), if
388: the GBR dominantly consists of photons whose energy has been
389: uplifted by ICS, as we proceed to show.
390:
391:
392: \section{The index of the GBR spectrum}
393:
394:
395: The current temperature, number density and mean
396: energy of the CMB are ${\rm
397: T_0=2.728}$
398: K, ${\rm n_0\approx 411~cm^{-3}}$, and
399: ${\rm \epsilon_0\approx} $ ${\rm 2.7\, kT_0\approx }$ ${\rm
400: 6.36\times10^{-10}~ MeV}$ (Mather et al.~1993; Fixsen et al.~1996).
401: The galactic starlight (SL) distribution is highly non-uniform,
402: its average energy is $\epsilon_\star\!\sim \!1$ eV.
403: Consider the ICS of high energy electrons on these radiations.
404: Assume the shape of the electron
405: flux, Eq.~(\ref{electrons}), observed at ${\rm E>10}$ GeV,
406: to be representative of the average galactic spectrum.
407: For the energy range of EGRET
408: the Thomson limit is accurate
409: even for ICS on SL, and the ${\rm e\gamma}$ cross section
410: is ${\rm \sigma_{_T}\approx 0.65\times 10^{-24}~cm^2}$.
411: The mean energy ${\rm E_\gamma}$ of the upscattered photons,
412: --or ${\rm \Delta E_e}$, the mean energy loss per collision-- is:
413: \begin{equation}
414: {\rm E_\gamma(\epsilon_i) \approx \Delta
415: E_e (\epsilon_i)\approx {4\over 3}\,\left({E_e\over m_e\,
416: c^2}\right)^2\,\epsilon_i} \, ,
417: \label{loss}
418: \end{equation}
419: with ${\rm \epsilon_i=\epsilon_0}$ or $\epsilon_\star$.
420:
421: The ICS photon spectrum originating in our galaxy is the sum of CMB
422: and SL contributions:
423: \begin{equation}
424: {\rm {dF_\gamma\over dE}=
425: {dF_\gamma^0\over dE}+
426: {dF_\gamma^\star\over dE}}\, ,
427: \label{ICSphotons}
428: \end{equation}
429: and is a function of the galactic latitude (b) and longitude (l) coordinates.
430: The ICS final-photon spectrum --a cumbersome
431: convolution (Felten \& Morrison 1966) of a CR power spectrum with a
432: photon thermal distribution-- can be approximated very simply.
433: Using again the index ``i'' to label the CMB and SL fluxes:
434: \begin{equation}
435: {\rm {dF^i_\gamma\over dE_\gamma}
436: \simeq {N_i(b,l)~\sigma_{_T}}~{dE^i_e\over dE_\gamma}~
437: \left[ {dF_e\over dE_e}
438: \right]_{E_e=
439: {E}_e^i}\,;~~~~ {E}_e^i\equiv m_ec^2
440: \sqrt{{3\, E_\gamma\over 4\,\epsilon_i}}},
441: \label{ICSCBR}
442: \end{equation}
443: where ${\rm E_e^i}$ is obtained from Eqs.~(\ref{loss}) by inverting
444: ${\rm E_\gamma(\epsilon_i)}$.
445: We postpone to Section 6 the discussion
446: of the model-dependent normalization factors
447: ${\rm N_\star(b,l)}$ and ${\rm N_0(b,l)}$: effective column densities resulting from the
448: convolution of the space distribution of CR electrons
449: with those of starlight and of the CMB.
450: Introducing the CR-electron flux of
451: Eq.~(\ref{electrons}), of the form ${\rm dF_e/dE=A\,[E/MeV]^{-\beta_e}}$, into Eqs.~(\ref{ICSCBR}), we obtain:
452: \begin{equation}
453: {\rm {dF^i_\gamma\over dE}=
454: {N_i(b,l)~\sigma_{_T}\,A\over 2}~
455: \left[{4\,\epsilon_i\,MeV\over 3\,m_e^2c^4}\right]^{{\beta_e-1\over 2}}
456: \,\left[{E\over MeV} \right]^{-{\beta_e+1\over 2}}
457: \propto
458: [E ]^{-2.10\pm 0.05}}\, .
459: \label{ICSphotpred}
460: \end{equation}
461: In the energy-range of EGRET, the CMB and
462: SL contributions have the same spectral index,
463: as do the small sunlight and external-galaxy contributions
464: discussed in Sections 8 and 9.
465:
466:
467: The photon spectral index of Eqs.~(\ref{ICSphotpred}),
468: which is related to that of the CR-electrons
469: through $\beta_\gamma=(\beta_e+1)/2$,
470: coincides with the measured one, Eq.~(\ref{photons}).
471: The electron spectrum of Eq.~(\ref{electrons})
472: describes the data in the range ${\rm E_e>5}$ GeV, so that
473: Eq.~(\ref{ICSphotpred}) should be valid above ${\rm E_\gamma\sim
474: 100}$ keV, the typical energy of photons up-scattered from the CMB.
475: At ${\rm E_\gamma>
476: 50}$ GeV, at the upper end of the EGRET data,
477: ${\rm \sigma_T}$ in the SL contribution should be replaced
478: by the complete Klein--Nishina cross section, implying a steepening
479: of the spectrum. The corresponding
480: effect for the CMB contribution is at energy above the EGRET energy range.
481:
482:
483: In deriving Eqs.~(\ref{ICSphotpred}), we have assumed that
484: the locally-measured slope of Eq.~(\ref{electrons})
485: is representative of the index of the spectrum of the electrons
486: suffering ICS to produce the GBR, wherever they may be.
487: The spectral index of the diffuse GBR observed by EGRET is
488: independent of direction, as shown in Fig.~3.
489: The statistical test for a flat distribution is surprisingly good:
490: ${\bar\chi^2\sim 0.5}$. This is encouraging support for our working
491: hypothesis of an electron spectrum with a universal shape, and of
492: a simple and dominant mechanism --ICS-- to generate the GBR.
493:
494:
495: \section{The index of the electron spectrum}
496:
497: To relate the spectra of CR electrons and protons,
498: we need an estimate of the protons' spectrum at their source.
499: A source spectrum
500: ${\rm dF^s/dE}$ with index ${\rm\beta_s\sim 2.2}$
501: is obtained from collisionless shock simulations (Bednarz \& Ostrowski 1998)
502: or analytical estimates of acceleration by relativistic jets (Dar 1998).
503: The CR spectrum of nuclei is modulated by their
504: residence time in the Galaxy, ${\rm \tau_{gal}(E)}$. For a steady source of
505: CRs the energy dependence of the observed flux is roughly that of
506: ${\rm \tau_{gal}\, dF^s/dE}$.
507: Observations of astrophysical and solar plasmas and of nuclear
508: abundances as functions of energy (e.g. Swordy et al.~1990) indicate
509: that
510: ${\rm \tau_{gal}(E)\propto E^{ -0.5\pm 0.1}}$,
511: explaining ${\rm \beta_1\sim \beta_s +0.5 \sim 2.7}$, as in Eq.~(\ref{protons}).
512:
513: Practically all CR acceleration mechanisms invoke
514: an ionized medium
515: that is swept by a moving magnetic field, such as would
516: be carried by the rarefied plasma in a supernova shell (Bhattacharjee
517: \& Sigl 2000)
518: or by a `plasmoid' of jetted ejecta (Dar \& Plaga 1999).
519: The magnetic field acts as a moving `mirror' that imparts the
520: same distribution in velocity, or Lorentz factor
521: ${\rm \gamma=E/m\,c^2}$, to all charged
522: particles. To the extent that particle-specific losses
523: (such as synchrotron radiation) can be neglected
524: at the acceleration stage, all source fluxes
525: have the same energy-dependence. For electrons
526: below the anticipated `electron's knee' at
527: ${\rm E_e=(m_e/m_p)\, E_{knee}}$$\sim 2$ TeV,
528: we expect ${\rm dF^s_e/dE\propto E^{-\beta_s}}$, with
529: $\beta_s\sim 2.2$.
530: Confinement effects preserve
531: this equality for ultrarelativistic electrons and protons: their
532: behaviour in a magnetic maze is the same.
533: But, unlike for hadrons, the `cooling' time of electrons
534: --that are significantly affected by the ambient radiation
535: and magnetic fields-- is shorter than their galactic
536: confinement time,
537: ${\rm \tau_{gal}(E)}$, above a relatively low energy. This implies
538: that the CR electron spectrum is modulated mainly by the ICS, and not
539: by the confinement time.
540:
541: Electrons lose energy not only by ICS on starlight and the CMB, but also by
542: synchrotron radiation on magnetic fields. All of these processes are
543: essentially the same: scattering off photons, either real or virtual.
544: The energy loss is governed by the
545: rate at which a single electron interacts with the ambient
546: electromagnetic fields, weighted by the corresponding average energy
547: density:
548: ${\rm P=\sigma_{_T}\, c\,[n_\star\epsilon_\star+n_0\epsilon_0+B^2/(8\pi)]}$.
549: Let ${\rm R_p}$ (an inverse time) be the production rate of CR
550: electrons, assumed to be constant (Berezinskii et al.~1990), and let
551: ${\rm dn^s_e/dE}$ be their source
552: number-density spectrum. The actual density ${\rm dn_e/dE}$
553: in an interval ${\rm dE}$ about ${\rm E}$ is
554: continuously replenished and depleted
555: by electrons whose energy is being degraded by interactions. This leads to
556: a steady-state situation in which production and losses are in balance. Using Eq.~(\ref{loss}) we obtain:
557: \begin{equation}
558: {\rm {4\over 3}\,{P\over (m_e\, c^2)^2}
559: \, {d\over dE} \left(E^2\, {dn_e\over
560: dE}\right)=
561: R_p\,{dn^s_e\over dE}}\, .
562: \label{balance}
563: \end{equation}
564:
565: For a relatively uniform galactic CR
566: electron density, Eq.~(\ref{balance}) also applies to the
567: local electron flux ${\rm dF_e\simeq (c/4\pi)dn_e}$.
568: Substitute the spectrum ${\rm
569: dn^s_e/dE\sim E^{-\beta_s}}$
570: into the flux version of Eq.~(\ref{balance}) to
571: obtain:
572: \begin{equation}
573: {\rm {dF_e\over dE}= {3\, m_e^2\, c^4\, R\over
574: 4\,(\beta_s-1)\,P}\; { dF^s_e\over
575: E\, dE}\propto E^{-(\beta_s+1)}}\, .
576: \label{electron}
577: \end{equation}
578: For electrons with ${\rm E_e<(m_e/m_p)\, E_{knee}}$
579: we deduced that
580: $\beta_s\sim 2.2~.$ Thus, $\beta_s+1=3.2$,
581: in agreement with the data:
582: Eq.~(\ref{electrons}) and Fig.~2.
583: Above the `electron's knee' at ${\rm E_e\sim 2}$ TeV
584: the spectrum should steepen up by $\Delta
585: \beta \simeq 0.25$, like that of CR hadrons (Dar 1998). The available
586: spectral measurements extend only to ${\rm E_e\leq 1.5}$ TeV.
587:
588: The energy density in the CMB is ${\rm n_0\epsilon_0=0.24}$ eV cm$^{-3}$,
589: coincidentally similar to that in starlight at our location:
590: ${\rm n_\star\epsilon_\star\sim 0.22}$ eV cm$^{-3}$. If the local CR
591: and magnetic energy densities are in equipartition,
592: ${\rm B^2/(8\pi)\sim 1}$ eV cm$^{-3}$, again in the same ballpark.
593: The cooling time of electrons in the ensemble of these fields is:
594: \begin{equation}
595: {\rm \tau_{_{cool}}(E)\simeq {3\,m_e^2\, c^4\over
596: 4 \, P\, E}
597: \simeq 0.22\times \left [{E\over GeV}\right]^{-1}~Gy}\, .
598: \label{coolingbis}
599: \end{equation}
600: The galactic escape time of GeV electrons, which should be similar to that of
601: CR protons
602: ${\rm \tau_{gal}(E)\propto E^{ -0.5\pm 0.1}}$ (Swordy et al.~1990),
603: has a weaker energy dependence than that of $\rm \tau_{_{cool}}$.
604: At sufficiently low energy, then, ${\rm \tau_{gal}<\tau_{_{cool}}}$,
605: and processes other than Compton- or synchrotron cooling
606: (such as Coulomb scattering, ionization losses and bremsstrahlung)
607: become relevant.
608: The slope of Eq.~(\ref{electron})
609: should change as the energy is lowered. The spectrum of
610: Fig.~2 shows such a change, but it occurs at ${\rm E<10~GeV}$,
611: a range in which local modulations would mask the effect.
612:
613:
614: \section{The scale height of CR electrons}
615:
616:
617: The radio emission of galaxies seen edge-on --interpreted as
618: synchrotron radiation
619: by electrons on their local magnetic field--
620: offers direct observational evidence for CR electrons well above
621: galactic disks (e.g. Duric et al.~1998). For the particularly well observed
622: case of NGC 5755, the exponential
623: scale height of the synchrotron radiation is ${\cal{O}} (4)$ kpc.
624: If the CRs and the magnetic field energy are in equilibrium, they
625: should have
626: similar distributions, and the exponential
627: scale height ${\rm h_e}$ of the electrons ought to
628: be roughly twice that of
629: the synchrotron intensity, which reflects the convolution of
630: the electron- and magnetic-field distributions. The
631: inferred value ${\rm h_e}\sim 8$
632: kpc for NGC 5755 may not be universal for spirals, since ${\rm h_e}$ is very
633: sensitive to the density and distribution of CR sources, gas and plasma
634: in each particular galaxy. Moreover, the magnetic field may be in equipartition with cosmic rays only where the interstellar plasma is dense enough. It is
635: quite possible for the CR electrons to be confined in a large
636: magnetic halo with a field much smaller than that in the disk.
637: For these reasons we must discuss the observations
638: of our own particular galaxy.
639:
640: Traditionally CR electrons and nuclei were assumed to have a distribution
641: that snugly fit that of the visible part of the Galaxy --where their
642: conventional sources lie-- implying a scale height above the plane of
643: the disk
644: of ${\cal{O}} (1)$ kpc (Broadbend et al.~1989). As the data and their
645: analysis
646: became more
647: elaborate, scale heights more than one order of magnitude larger were
648: discussed (e.g. Strong et al.~1998). Since electrons lose energy to the ambient
649: radiation close to their sources, which have traditionally been
650: located in the
651: disk, not very well understood
652: CR-reacceleration phenomena have had to be
653: invoked (e.g. Seo \& Ptuskin 1994).
654: Even with reacceleration, a conventional distribution of cosmic-ray
655: sources fails to describe the observed GBR (Strong \& Moskalenko 1998).
656:
657: Over the years,
658: Moskalenko, Strong and their collaborators have developed
659: what is presumably the most
660: elaborate and detailed understanding of the CR, radio and $\gamma$
661: observations of our galaxy
662: (Moskalenko et al.~1998; Moskalenko and Strong, 2000;
663: Strong and Moskalenko, 1998; Strong et al.~1997;
664: Strong et al.~1998). A crucial parameter in their models is the scale ${\rm z_h}$
665: of the CR distribution orthogonal to the galactic plane, defined as
666: the height above which CRs freely escape, as in a leaky-box model.
667: Strong \& Moskalenko (1998) conclude
668: that ${\rm z_h}$ lies between 4 and 12 kpc. The limits are based on the
669: comparison of the ${\rm ^{10}Be/^9Be}$ ratio observed by
670: Ulysses (Connell 1998)
671: with model predictions as a function of ${\rm z_h}$, being all other
672: parameters
673: fixed at their adopted values. The dependence of the ${\rm ^{10}Be/^9Be}$
674: ratio on ${\rm z_h}$, shown in Fig.~9 of Strong \& Moskalenko (1998)
675: and reproduced here as Fig.~5, is very weak for
676: ${\rm z_h>10}$ kpc. At ${\rm z_h=20}$ kpc, the prediction would
677: be only some 1.3 standard
678: deviations below the Ulysses central value, and even ${\rm z_h=40}$ would
679: be viable: the average of all previous and somewhat less precise
680: observations,
681: compiled in Lukasiak et al.~(1994) and shown in
682: Fig.~5a, would be in agreement with ${\rm z_h=}$
683: 20 or 40 kpc. For all these reasons and
684: the ones stated in the introduction, we shall not refrain from considering
685: scale heights above the 12 kpc upper limit quoted by Strong \&
686: Moskalenko (1998).
687:
688: \section{The CBB and SL contributions to the GBR}
689:
690: The spectral index of the GBR, derived in Section 4,
691: is independent of the details of the spatial distribution
692: of starlight. We have argued that the EGRET GBR data
693: support the simple hypothesis of an electron spectral
694: index that is independent of location. The predicted
695: GBR index is then also independent of the magnitude of the
696: electron spectrum as a function of position.
697: In this section we use a simplified model of the electron and starlight
698: distributions to compute the magnitude and angular
699: dependence of the CMB and SL contributions to the GBR.
700:
701:
702: We adopt ${\rm h_e=20}$ kpc (a value obtained from a
703: rough fit of our results to the angularly-averaged fluence of
704: the GBR) for the Gaussian
705: scale height of the CR electron distribution
706: of our galaxy in the direction perpendicular to the galactic plane.
707: For the
708: distribution in $\rho$ --the radial coordinate orthogonal to the
709: galactic axis-- we
710: adopt a Gaussian scale height $\rm{\rho_e=35}$ kpc; the results are
711: quite insensitive to this
712: parameter. The EGRET GBR data are not precise enough
713: to be ``invertible'', that is, for the
714: actual high-latitude CR-electron distribution (Gaussian, exponential
715: or otherwise) to be disentangled; a fact to be rediscussed anon,
716: in view of our results.
717: The distance of the solar system to the galactic centre
718: is ${\rm d_{_\odot}\simeq 8.5}$ kpc. The factor
719: ${\rm N_0(\theta,\phi)}$ in Eq.~(\ref{ICSCBR}), which describes the
720: angular dependence
721: of the GBR photons due to ICS on the (uniformly distributed) CMB, is:
722: \begin{eqnarray}
723: &&{\rm N_0(b,l)= \int_0^\infty \, dr \;n_0\;
724: Exp\left[+\left({d_{_\odot} \over \rho_e}\right)^2\right]
725: Exp\left[-\,\left({h(r,b)\over h_e}\right)^2
726: -\left({\rho(r,b,l)\over \rho_e}\right)^2\,\right]}\, ,\nonumber \\
727: &&{\rm h(r,b)\equiv r\,sin\,b }\, ,\nonumber \\
728: &&{\rm \rho(r,b,l)\equiv
729: \left([r\,\cos\,(b)\,\cos\,(l)-d_{_\odot}]^2+[r\,\cos\,(b)\,\sin\,(l)]^2\right)^{1/2}}\, ,
730: \label{CBRangle}
731: \end{eqnarray}
732: where r is the distance in the direction along the line of sight.
733:
734: It is difficult to model in detail the contributionn from ICS on
735: starlight (Hunter et al.~1997, Sreekumar et al.~1998). But we are only
736: concerned with this light
737: at high galactic latitudes, since the diffuse GBR of interest to us is that
738: measured by EGRET by masking the galactic plane and
739: centre.
740: We make a coarse estimate by approximating the Galaxy's starlight
741: as that produced by a source at its centre with the galactic
742: luminosity ${\rm L_\star}=2.3\times 10^{10}$ ${\rm L_{_\odot}}$
743: $\simeq 5.5~10^{55}$ eV s$^{-1}$ (Pritchet \& van den Bergh 1999).
744: The starlight
745: contribution in Eq.~(\ref{ICSphotons}) is then of the same form as
746: Eq.~(\ref{CBRangle}),
747: with ${\rm N_0}$ traded for ${\rm N_\star}$ by the substitution:
748: \begin{eqnarray}
749: {\rm n_0\rightarrow {L_\star\over 4\,\pi\,c\,\epsilon_\star}\;\;
750: {1\over( r^2-2\,r\,d_{_\odot}\,\cos\,(b)\,\cos\,(l)+d_{_\odot}^2)}}\, .
751: \label{Lightangle}
752: \end{eqnarray}
753:
754: For the CMB and starlight contributions to the GBR,
755: averaged over the EGRET unmasked domain, we obtain, by integration of
756: Eqs.~(\ref{ICSphotons}), (\ref{ICSCBR}), (\ref{CBRangle}), (\ref{Lightangle}):
757: \begin{equation}
758: {\rm {dF_\gamma\over dE}
759: \simeq (2.41\pm 0.55)\times\! 10^{-3}\left[
760: E\over MeV \right]^{-2.10\pm 0.05}
761: cm^{-2}s^{-1}sr^{-1}MeV^{-1}}.
762: \label{ICSphotons2}
763: \end{equation}
764: For scale heights ${\rm h_e}$ and $\rho_{\rm e}$ similar to
765: the ones adopted (20 and 35 kpc,
766: respectively), the CMB and SL contributions are comparable
767: in magnitude, the first scales approximately linearly
768: with ${\rm h_e}$ while the second is rather
769: insensitive to this parameter. The contribution to the CMB
770: from sunlight and external galaxies, discussed in
771: Section 8 and 9, adds corrections of 6\% and
772: $\sim \! \!10\%$ (respectively) to Eq.~(\ref{ICSphotons2}),
773: the total result is shown in Fig.~2.
774: The fitted value of ${\rm h_e}$ is imprecise: the starlight to CMB ratio
775: is proportional to $\epsilon_\star/\epsilon_0$ raised to a
776: very poorly determined power, $0.10\pm 0.05$.
777:
778: We can use our assumed Gaussian distribution of
779: electrons in a halo, with vertical and radial
780: scale heights ${\rm h_e}$ and ${\rm \rho_e}$, to compute
781: the diffuse $\gamma$-ray luminosity of our galaxy,
782: which in our model is dominated by ICS on CMB and SL photons.
783: Using
784: Eqs.~(\ref{ICSphotons}), (\ref{ICSCBR}), (\ref{CBRangle}), (\ref{Lightangle})
785: we obtain, for the luminosity in
786: $\gamma$-rays of energy above E:
787: \begin{eqnarray}
788: &&{\rm L_\gamma (>E) \simeq L^0_\gamma (>E)
789: +L^\star_\gamma(>E)}\, ,\nonumber \\
790: &&{\rm L^0_\gamma (>E)=1.31\times 10^{40}
791: \left[{\rho_e\over 35~kpc}\right]^2
792: \left[{h_e\over 20~kpc}\right]
793: \left[{E\over MeV}\right]^{-0.10\pm 0.05}~erg/s} ,\nonumber \\
794: &&{\rm L^\star_\gamma (>E)= 3.56\times 10^{39}
795: \left[{h_e\over 20~kpc}\right]\, \left[{1\over 2\, u}\, ln{1+u\over
796: 1-u}\right] \left[{E\over MeV}\right]^{-0.10\pm 0.05}~erg/s},
797: \label{radiation}
798: \end{eqnarray}
799: where ${\rm u\equiv \sqrt{1-h_e^2/\rho_e^2}}$.
800: A future $\gamma$-ray telescope, such as GLAST, could possibly
801: see the corresponding glow of Andromeda's halo.
802:
803: \section{Sunlight contribution to the local GBR}
804:
805: We are only at a distance ${\rm l_{_\odot}=1.5\times 10^{13}~cm}$
806: from the sun. This
807: entails a small but non-negligible
808: contribution to the locally-observed GBR, resulting from
809: ICS off photons in the heliosphere. The corresponding photon
810: flux is described by
811: Eq.~(\ref{ICSphotpred}), with the substitution of $\epsilon_i$ by the
812: mean energy
813: ${\rm \epsilon_{_\odot}\approx 1.35~eV}$ of solar photons, and of ${\rm N_i}$
814: by ${\rm N_{_\odot}}$,
815: the solar-photon column density along the line of sight. Let
816: $\theta_{_\odot}$ be the angle between the line of sight and the direction
817: to the sun. Then:
818: \begin{equation}
819: {\rm N_{_\odot}(cos\theta_{_\odot})=
820: {L_{_\odot} \over 4\,\pi\,c\,l_{_\odot}\epsilon_{_\odot}}\,
821: \left( {\pi-\theta_{_\odot}\over \sin\theta_{_\odot} } \right) }\, .
822: \label{sun}
823: \end{equation}
824: For a uniform ${\rm cos\,\theta_{_\odot}}$ distribution
825: during the EGRET data taking, the average column density is
826: ${\rm \overline N_{_\odot}=\pi
827: \,L_{_\odot}/(16\,c\,l_{_\odot}\epsilon_{_\odot})}$, resulting
828: in a sunlight-induced GBR flux:
829: \begin{equation}
830: {\rm {dF_\gamma^\odot \over dE}\approx 1.32\times 10^{-4}
831: \left[{E\over MeV}\right]~cm^{-2}s^{-1}sr^{-1}MeV^{-1}}.
832: \label{suntot}
833: \end{equation}
834: This contribution is roughly 6\% of our galaxy's
835: result, Eq.~(\ref{ICSphotons2}).
836: At ${\rm E_\gamma>
837: 75}$ GeV, the spectrum of Eq.~(\ref{suntot}) should
838: steepen, since ICS should then be described by the
839: Klein--Nishina cross section, and not by its low energy
840: Thomson limit.
841:
842:
843:
844: \section{Extragalactic contribution to the GBR}
845:
846: To estimate this contribution, some concepts and numbers need to be recalled.
847: Hubble's constan' is ${\rm H_0=100~h~km~s^{-1}Mpc^{-1}}$,
848: with ${\rm h\sim 0.65}$;
849: ${\rm \Omega_m}$ and $\Omega_\Lambda$ are matter and vacuum
850: cosmic densities in critical units:
851: ${\rm \Omega\equiv\Omega_m+ \Omega_\Lambda}$;
852: $y\equiv 1+z$ is the redshift factor.
853: In a Friedman model, the time to redshift relation is
854: ${\rm dy/dt=-H_0\, f(y)\, y}$, with
855: ${\rm f(y)\equiv [(1-\Omega)\, y^2+\Omega_m\, y^3+\Omega_\Lambda]^{1/2}}$.
856: The luminosity density of the local universe (Ellis 1997) is
857: ${\rm \rho_{_L}=(2.0\pm 0.4)\times 10^8\, h\, L_{_\odot}\, Mpc^{-3}}$.
858: The
859: combination ${\rm \rho_{_L}/L_\star}$ provides an estimate of the average
860: number density of `Milky-Way-equivalent' galaxies. If the main sources
861: of CRs are young supernova remnants or gamma-ray bursts,
862: the CR production rate ought to be proportional (e.g. Wijers et al.~1997)
863: to the star formation rate ${\rm R_{SFR}[y]}$,
864: recently measured up to redshift
865: ${\rm z\simeq 4.5}$ (Steidel et al.~1998).
866:
867:
868: The energy of CMB photons up-scattered by electrons at `epoch y'
869: is proportional
870: to ${\rm T(y)=y~T_0}$ and it is subsequently redshifted by the same factor;
871: hence the spectra from distant galaxies should have
872: the same energy dependence as from our galaxy. The situation for
873: SL photons is more complicated. Young galaxies are bluer than older
874: ones, but this effect is overcompensated by the expansion redshift
875: from a relatively low y, onwards. Yet, at the energies observed by EGRET,
876: and for the redshift values of $\cal{O}$(1) that dominate
877: the extragalactic contribution, all these
878: blue- and red-shifts simply relocate the photon energy,
879: while roughly maintaining the slope of the spectrum.
880: For the sum of all galaxies, we estimate:
881: \begin{equation}
882: {\rm {dF_\gamma^{^{EG}}\over dE}\sim
883: {1\over 4\,\pi}~{dL_\gamma\over E\, dE}~{\rho_{_L}\over L_*}
884: ~{c\over H_{0}}\,
885: \int_1 {R_{SFR}(y)\over R_{SFR}(0)}~{y\over f(y)}~{dy\over y^3} },
886: \label{allgals}
887: \end{equation}
888: where ${\rm dL_\gamma/dE}$ is to be obtained
889: from the luminosity of a Milky-Way-like galaxy, Eq.~(\ref{radiation}).
890: For ${\rm R_{SFR}[y]}$ we interpolate the summary
891: values of Steidel et al. (1998). In writing Eq.~(\ref{allgals})
892: we have ignored the fact that, above
893: ${\rm E\sim 10}$ GeV,
894: absorption by $e^+e^-$ production on the IR-to-UV background becomes
895: relevant (Salamon \& Stecker 1998),
896: so that the extragalactic contribution should be quenched.
897:
898: For ${\rm \Omega=\Omega_m=1}$ the value of the
899: integral in Eq.~(\ref{allgals}) is $\sim 0.82$; it increases to
900: $\sim 1.08$ for a currently more fashionable universe with
901: ${\rm \Omega=1}$, ${\rm \Omega_\Lambda=0.7}$,
902: ${\rm \Omega_m=0.3}$. For the latter case, the result is:
903: \begin{equation}
904: {\rm {dF^{^{EG}}_\gamma\over dE}
905: = 2.48 \times\! 10^{-4} \left[
906: E \over MeV \right]^{-2.10\pm 0.05}
907: cm^{-2}s^{-1}sr^{-1}MeV^{-1}}\, ,
908: \label{allgals2}
909: \end{equation}
910: roughly $10\%$ of our galaxy's angularly-averaged result,
911: Eq.~(\ref{ICSphotons2}).
912:
913: \section{Detailed comparison with the EGRET data}
914:
915:
916: Our predictions for the magnitude of the GBR
917: and its directional dependence on b and l are shown in
918: Figs.~5 and 6. In Fig.~5 we display separately the contributions
919: from ICS off CMB and SL photons in our galaxy, as well as the uniformly
920: distributed sunlight and extragalactic components.
921: In Fig.~7 we compare
922: the total GBR flux:
923: \begin{equation}
924: {\rm {dF_\gamma\over dE}=
925: {dF_\gamma^0\over dE}+
926: {dF_\gamma^\star\over dE}+
927: {dF_\gamma^\odot \over dE}+
928: {dF_\gamma^{^{EG}}\over dE}
929: }\, ,
930: \label{total}
931: \end{equation}
932: obtained by summing Eqs.~(\ref{ICSphotpred}),
933: Eq.~(\ref{suntot}) and Eq.~(\ref{ICSphotons2}),
934: with the EGRET data.
935: Our result is a satisfactory fit to the observed
936: magnitude and angular trend of the GBR
937: (${\bar\chi^2=0.98}$), a vast improvement over the result for a
938: constant (extragalactic) ansatz, for which ${\bar\chi^2=2.6}$.
939: Although this agreement would be more meaningful, had we used a
940: more realistic model of starlight,
941: a more careful treatment may be premature, for the EGRET error
942: bars are large enough to accommodate considerable variations in
943: the input modelling. In a previous analysis (Dar et al.~1999), for instance,
944: we obtained a similarly good fit with an assumed
945: constant-density, spherical CR-electron halo of radius 25 kpc,
946: for which the results have the advantage of being simple
947: analytical functions.
948:
949: We have neglected various putative extragalactic contributions to
950: the GBR. Blazars,
951: because of their beamed emission, may not be very relevant.
952: But CR electrons injected directly into
953: intergalactic space by active galactic nuclei,
954: radio galaxies or gamma ray bursters, may
955: give rise to a contribution of comparable magnitude and shape
956: to that of the CR electrons in external galaxies.
957: These or other potential sources of GBR photons
958: may imply that our parameters ${\rm h_e}$ and ${\rm \rho_e}$
959: have been overestimated. But this effect cannot be very large,
960: given our success at describing the non-trivial angular dependence
961: of the EGRET data.
962:
963:
964: \section{Conclusions and predictions}
965:
966: We have presented a simple understanding of the relation
967: between the spectral indices of cosmic-ray protons, electrons
968: and the GBR. Accepting the possibility that the CR-electron
969: distribution in our galaxy may have a scale height larger
970: than conventionally believed, we have also argued that
971: the bulk of the GBR could originate in our own galaxy.
972: Our modelling is extremely simplistic, but quite successful.
973:
974: The predictions specific to our scenario are:
975: \begin{itemize}
976: %
977: \item{} The GBR should reflect the
978: asymmetry of our off-centre position in the Galaxy.
979: %
980: \item{} The halo of Andromeda
981: should shine in gamma rays above a few MeV, with a luminosity
982: comparable to that in Eq.~(\ref{radiation}).
983: Likewise, very nearby star-burst Galaxies, such as M82,
984: and radio galaxies with large CR production rates, such as Cygnus A,
985: may be visible in gamma rays.
986: %
987: \item{} If the CR-proton and electron acceleration mechanisms are the same,
988: the existence of a knee in the observed proton spectrum translates
989: into a related result for the
990: power index ${\rm \beta_e}$ of the electron spectrum, which should
991: steepen above
992: ${\rm E\approx 1.6}$ TeV by $\Delta\beta\sim 1/4$. %
993: %
994: \item{} The GBR spectrum should not have the sharp cutoff,
995: above ${\rm E\sim 100}$ GeV,
996: expected (Salamon \& Stecker 1998) for cosmological sources.
997: But it should nonetheless steepen around 10--100 GeV, because of
998: the anticipated ``knee'' in the electron spectrum and of the
999: energy-dependence of the Klein-Nishina cross section.
1000:
1001: \end{itemize}
1002:
1003:
1004: These features of our scenario should be testable
1005: when the next generation of cosmic-ray and
1006: $\gamma$-ray satellites (AMS-02 and GLAST) are
1007: operational, hopefully by 2005. In spite of their
1008: maturity, cosmic-ray physics and $\gamma$-ray
1009: astrophysics are still young, and thriving.
1010:
1011: ACKNOWLEDGEMENTS\\
1012: We are indebted to G. Bignami, S. Dado and G. Raffelt
1013: for discussions and to I. Moskalenko and A. Strong
1014: for permission to reproduce their results in
1015: our Fig. 5.
1016:
1017: \newpage
1018: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1019: %\begin{thebibliography}{99}
1020: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1021: %\begin{thebibliography}{99}
1022: %\bibitem{MSdistrib} MS distrib of CR sources
1023:
1024: \noindent
1025: REFERENCES\\
1026: Barwick S. W. et al., 1998, ApJ, 498, 779 \\
1027: Bednarz J., Ostrowski M., 1998, PRL, 80, 3911\\
1028: Berezinskii V. S. et al., 1990, {\it Astrophysics of cosmic rays} (North
1029: Holland, Amsterdam, 1990)\\
1030: Bhattacharjee P., Sigl G., 2000, Phys. Rep., 327, 109\\
1031: Bignami G. et al., 1979, ApJ, 232, 649\\
1032: Bird D. J. et al., 1995, ApJ, 441, 144.\\
1033: Broadbend A., Haslam C. G. T., Osborne, J. L., 1989, MNRAS, 237, 381 \\
1034: %Carlberg et al 1997 ???\\
1035: Chiang J., Mukerjee R., 1998, ApJ, 496, 772\\
1036: Cohen A., De R\'ujula A., Glashow S. L., 1998, ApJ, 495, 539\\
1037: Connell J. J., 1998, ApJ, 501, L59\\
1038: Dar A., 1998, astro-ph/9809163, in {\it Proceedings of the Rencontres de la
1039: Vall\'ee d'Aoste,} 1998 (ed. M. Greco), Frascati Physics Series, INFN
1040: Pubs, page 23\\
1041: Dar A., De R\'ujula A., Antoniou N., 1999, astro-ph/9901004\\
1042: Dar A., Plaga R., 1999, A\&A, 349, 259\\
1043: Dixon D. D. et al., 1998, New Astron. 3, 539\\
1044: Duric N., Irwin J., Bloemen H., 1998, A\& A, 331, 428\\
1045: Ellis R. S., 1997, ARA\&A, 35, 389 \\
1046: Evenson P., Meyers P., 1984, J. Geophys. Res., 89 A5, 2647 \\
1047: Felten J. E., Morrison P., 1966, ApJ, 146, 686\\
1048: Ferrando P. et al., 1996, A\&A 316, 528 \\
1049: Fixsen D.J. et al., 1996, ApJ, 473, 576\\
1050: Gnedin N. Y., Ostriker J. P., 1992, ApJ, 400, 1 \\
1051: Golden R. L. et al., 1984, ApJ, 287, 622\\
1052: Golden R. L., et al., 1994, ApJ, 436, 739 \\
1053: Hartmann D. H., 1995, ApJ, 447, 646\\
1054: Hawking S. W., 1977, Scientific American, 236, 34\\
1055: Hunter S. D. et al., 1997, ApJ, 481, 205\\
1056: Kazanas D., Protheroe J. P., 1983, Nature, 302, 228\\
1057: %Lin et al 1996 ??? \\
1058: Lukasiak A. et al., 1994, ApJ, 423, 426\\
1059: Mather J. C. et al., 1993, ApJ, 432, L15\\
1060: Moskalenko I. V., Strong A. W., Reimer O., 1998, astro-ph/9811221\\
1061: Moskalenko I. V., Strong A. W., 2000, ApJ, 528, 357\\
1062: Nishimura J. et al., 1980, ApJ, 238, 394 \\
1063: Page D. N., Hawking S. W., 1976, ApJ, 206, 1 \\
1064: Pohl M., Esposito J. A., 1998, ApJ, 507, 327\\
1065: Prince T. A., 1979, ApJ, 227, 676\\
1066: Pritchet C. J., van den Bergh S., 1999, AJ, 118, 833\\
1067: Rudaz S., Stecker F. W., 1991, ApJ, 368, 40\\
1068: Salamon M. H., Stecker F. W., 1998, ApJ, 493, 547\\
1069: Seo E. S., Ptuskin V. S., 1994, ApJ, 431, 705\\
1070: Silk J., Srednicki M., 1984, PRL, 53, 264\\
1071: %Small et al 1998 ??? \\
1072: Sreekumar P. et al., 1998, ApJ, 494, 523\\
1073: Stecker F. W., Salamon M. H., 1996, ApJ, 464, 600\\
1074: Stecker F. W., Morgan D. L., Bredekamp J., 1971, PRL, 27, 1469\\
1075: Steidel C.C. et al., 1999, ApJ, 519, 1\\
1076: Strong A., Moskalenko I. V., 1998, ApJ, 509, 212\\
1077: Strong A. W. et al., 1997, in {\it Proceedings of the 4th Compton
1078: Symposium}, AIP, 410, 1198\\
1079: Strong A. W., Moskalenko I. V., Reimer O., 1998, astro-ph/9811296\\
1080: Swordy S. P. et al., 1990, ApJ, 330, 625 \\
1081: Takeda M. et al., 1998, PRL, 81, 1163\\
1082: Tang K. K., 1984, ApJ, 278, 881 \\
1083: Thompson D. J., Fichtel C. E., 1982, A\&A, 109, 352\\
1084: Webber W. R., 1997, Sp. Sci. Rev., 81, 107 \\
1085: Wiebel-Sooth B., Biermann, P. L., 1998, Landolt-B\"ornstein,
1086: (Springer Verlag, Heidelberg 1998, in press) \\
1087: Wijers R. A. M. J. et al., 1997, MNRAS, 294, L13 \\
1088:
1089:
1090: %\bibitem{H?} F. Halzen {\it et al.~,} Astropart. Phys.
1091: %{\bf 3}, 151 (1995).
1092: %J.W. Elbert and P. Sommers, Astroph. J. {\bf 441}, 151 (1995).
1093: %\bibitem{Lee} S. Lee, Phys. Rev. {\bf D58}, 043004 (1997).
1094: %\bibitem{GZK} K. Greisen, Phys. Rev. Lett. {\bf 16}, 748 (1966).
1095: %G.T. Zatsepin and V.A. Kuz'min, JETP Lett. {\bf 4}, 78 (1966).
1096:
1097:
1098: %\bibitem{Simp} See, e.g., J.A. Simpson and J.J. Connell,
1099: % Astroph. J. {\bf 497}, L88 (1998).
1100:
1101:
1102: %\bibitem{Barb}
1103: %R.L. Golden {\it et al.~,} Astroph. J. {\bf 436}, 739 (1994).
1104: %G. Barbiellini {\it et al.~,} Astron. and Astroph. {\bf 309}, L15 (1996).
1105: %S.W. Barwick {\it et al.~,} Astroph. J. {\bf 498}, 779 (1998).
1106: %S. Coutu {\it et al.~}, to appear in Astropar. Phys.
1107: %(astro-ph/99002162).
1108:
1109: %\bibitem{GS} V.L. Ginzburg and S. I. Syrovatskii, {\it
1110: %The Origin of Cosmic Rays} (New York: MacMillan 1964).
1111:
1112: %\bibitem{BB} M.C. Begelman and R.D. Blandford, Rev. Mod. Phys.
1113: %{\bf 56}, 255 (1984).
1114:
1115: %\bibitem{BR} P. Biermann and J. Rachen, Astron. Astroph.
1116: %{\bf 272}, 161 (1993).
1117:
1118:
1119:
1120: %\end{thebibliography}
1121:
1122: \newpage
1123:
1124:
1125: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%2
1126: \newpage
1127: \begin{figure}
1128: %\plotone{gbrfig1.eps}
1129: \begin{center}
1130: \vspace*{-1.6cm}
1131: \hspace*{-1cm}
1132: \epsfig{file=gbrfig1.eps,width=15cm}
1133: \caption{Comparison between the spectrum of
1134: the GBR, measured by EGRET (Sreekumar et al.~1998),
1135: and the prediction for ICS of starlight and the CMB by CR
1136: electrons. The slope is our central prediction, the normalization
1137: is the one obtained for ${\rm h_e}= 20$ kpc, ${\rm \rho_e}= 35$ kpc.}
1138: \vspace*{-0.5cm}
1139: \end{center}
1140: \end{figure}
1141:
1142: %%%%%%%%%%%%%%%%%%%%%%%%%%%%1
1143: \begin{figure}
1144: \begin{center}
1145: \vspace*{-1.6cm}
1146: \hspace*{-1cm}
1147: \epsfig{file=gbr0.eps,width=15cm}
1148: %\vspace*{-14.6cm}
1149: \caption{The primary cosmic-ray electron spectrum
1150: (Evenson \&
1151: Meyers 1984; Golden et al.~1994; Ferrando et al.~1996) as
1152: measured by Prince 1979 [crosses]; Nishimura et al.~1980 [squares];
1153: Tang 1984 [circles]; Golden et al.~1984 [triangles]; Barwick et
1154: al.~1998 [stars]. The slope is the prediction, the
1155: magnitude is normalized to the data.}
1156: \vspace*{-0.5cm}
1157: \end{center}
1158: \end{figure}
1159:
1160:
1161: %%%%%%%%%%%%%%%%%%%%%%%%%3
1162: %____________________________________________________________
1163: \begin{figure}[t]
1164: \begin{tabular}{cc}
1165: \hskip -0.5truecm
1166: \epsfig{file=pfign1.eps,width=8cm} &
1167: \hskip -0.5truecm
1168: \epsfig{file=pfign2.eps,width=8cm}
1169: \vspace{-0.5cm}\\
1170: %\hskip 0.2truecm
1171: {\small (a)} &
1172: \hskip -0.2truecm
1173: {\small (b)} \\
1174: \epsfig{file=pfign3.eps,width=8cm} &
1175: \hskip -0.5truecm
1176: \epsfig{file=pfign4.eps,width=8cm}
1177: \vspace{-0.5cm}\\
1178: \hskip 0.2truecm
1179: {\small (c)} &
1180: \hskip -0.20truecm
1181: {\small (d)}
1182: \end{tabular}
1183: \caption{{
1184: EGRET data on the GBR spectral
1185: index as a function of $\theta$
1186: the angle away from the direction of the galactic center.
1187: The line is the predicted spectral index.
1188: The various plots correspond to the individual half-hemispheres.
1189: (a)
1190: ${\rm b> 0}$, ${\rm l> 0}$.
1191: (b)
1192: ${\rm b> 0}$, ${\rm l< 0}$.
1193: (c)
1194: ${\rm b< 0}$, ${\rm l> 0}$.
1195: (d)
1196: ${\rm b< 0}$, ${\rm l< 0}$.}}
1197: \label{fig:indexes}
1198: \end{figure}
1199: %__________________________________________________________
1200:
1201: %____________________________________________________________
1202: \begin{figure}[t]
1203: \begin{tabular}{cc}
1204: \hskip -0.5truecm
1205: %\epsfig{file=yfign1clean.eps,width=7cm} &
1206: \epsfig{file=yfignew1.eps,width=7cm} &
1207: \hskip 0.5truecm
1208: %\epsfig{file=yfign2clean.eps,width=7cm}
1209: \epsfig{file=yfignew2.eps,width=7cm}
1210: \vspace{-0.5cm}\\
1211: %\hskip 0.2truecm
1212: {\small (a)} &
1213: \hskip 1.2truecm
1214: {\small (b)} \\
1215: \vspace{-0.9cm}\\
1216: \hskip -0.5truecm
1217: %\epsfig{file=yfign3clean.eps,width=7cm} &
1218: \epsfig{file=yfignew3.eps,width=7cm} &
1219: \hskip 0.5truecm
1220: %\epsfig{file=yfign4clean.eps,width=7cm}
1221: \epsfig{file=yfignew4.eps,width=7cm}
1222: \vspace{-0.5cm}\\
1223: \hskip 0.2truecm
1224: {\small (c)} &
1225: \hskip 1.2truecm
1226: {\small (d)}
1227: \end{tabular}
1228: \caption{{EGRET data, organized as in Fig.~3,
1229: for the dependence on $\theta$ of the GBR intensity
1230: above 100 MeV. }}
1231: \label{fig:intensities1}
1232: \end{figure}
1233: %__________________________________________________________
1234:
1235: %%%%%%%%%%%%%%%%%%%%%%%%%4
1236: %____________________________________________________________
1237: \begin{figure}[t]
1238: \begin{tabular}{cc}
1239: \hskip -0.5truecm
1240: \epsfig{file=fig9a.ps,width=8cm} &
1241: \hskip -0.5truecm
1242: \epsfig{file=fig9b.ps,width=8cm}
1243: \vspace{-0.5cm}\\
1244: %\hskip 0.2truecm
1245: %{\small (a)} &
1246: %\hskip 0.2truecm
1247: %{\small (b)}
1248: \end{tabular}
1249: \vspace{0.5 cm}
1250: \caption{{${\rm ^{10}Be/ ^9Be}$ ratio for the diffusive reacceleration models
1251: of Strong and Moskalenko (1998). (a) As a function of energy for
1252: ${\rm z_h=1}$, 2, 3, 4, 5, 10, 15 and 20 kpc. (b) As a function of
1253: ${\rm z_h}$
1254: at 525 MeV/nucleon, the mean interstellar value for the Ulysses data,
1255: whose 1$\sigma$ limits are the dashed lines.
1256: The data points in (a) are from Lukasiak el al.~1994
1257: (square, Voyagers 1,2; open circle, IMP 7/8; triangle, ISEE 3) and Connell
1258: 1998 (filled circle, Ulysses).}}
1259: \label{fig:moskastrong}
1260: \end{figure}
1261: %________________________________________________________
1262:
1263: \begin{figure}
1264: \begin{center}
1265: \vspace*{-1.6cm}
1266: \hspace*{-1cm}
1267: \epsfig{file=partsall1.eps,width=6cm}
1268: %\vspace*{-14.6cm}
1269: \caption{Contributions to the GBR flux above 100 MeV
1270: as functions of longitude l, at fixed latitude b, from ICS
1271: of starlight (dotted), and CMB in our galaxy (dashed); from
1272: the total ICS from external galaxies (continuous), and
1273: from sunlight (dot-dashed). The
1274: vertical scale is
1275: $10^4$ times the number of photons/${[\rm cm^2\,s\,sr]}$.
1276: The results are
1277: for ${\rm h_e=20}$ kpc, ${\rm \rho_e=35}$ kpc.}
1278: \vspace*{-0.5cm}
1279: \end{center}
1280: \end{figure}
1281:
1282: \begin{figure}
1283: \begin{center}
1284: \vspace*{-1.6cm}
1285: \hspace*{-1cm}
1286: \epsfig{file=allb1.eps,width=12cm}
1287: %\vspace*{-14.6cm}
1288: \caption{The flux of GBR photons above 100 MeV: comparison
1289: between EGRET data and our model
1290: for ${\rm h_e=20}$ kpc, ${\rm \rho_e=35}$ kpc, as functions
1291: of latitude at various fixed longitudes. The grey domain
1292: is EGRET's mask.}
1293: \vspace*{-0.5cm}
1294: \end{center}
1295: \end{figure}
1296:
1297: \end{document}
1298:
1299:
1300:
1301: