Very high energy gamma rays from the composite SNR G0.9+0.1

0.18 MB·English

by  H. E. S. S. Collaboration

Astronomy&Astrophysicsmanuscriptno.Gl153 February2,20108
(DOI:willbeinsertedbyhandlater)
Very high energy gamma rays from the composite SNR G0.9+0.1
F.Aharonian1,A.G.Akhperjanian2,K.-M.Aye3,A.R.Bazer-Bachi4,M.Beilicke5,W.Benbow1,D.Berge1,
P.Berghaus6 ⋆,K.Bernlo¨hr1,7,C.Boisson8,O.Bolz1,C.Borgmeier7,I.Braun1,F.Breitling7,A.M.Brown3,
J.BussonsGordo9,P.M.Chadwick3,L.-M.Chounet10,R.Cornils5,L.Costamante1,20,B.Degrange10,
A.Djannati-Ata¨i6,L.O’C.Drury11,G.Dubus10,T.Ergin7,P.Espigat6,F.Feinstein9,P.Fleury10,G.Fontaine10,
S.Funk1,Y.A.Gallant9,B.Giebels10,S.Gillessen1,P.Goret12,C.Hadjichristidis3,M.Hauser13,G.Heinzelmann5,
5 G.Henri14,G.Hermann1,J.A.Hinton1,W.Hofmann1,M.Holleran15,D.Horns1,O.C.deJager15,I.Jung1,13 ⋆⋆,
0
B.Khe´lifi1,Nu.Komin7,A.Konopelko1,7,I.J.Latham3,R.LeGallou3,A.Lemie`re6,M.Lemoine10,N.Leroy10,
0
2 T.Lohse7,A.Marcowith4,C.Masterson1,20,T.J.L.McComb3,M.deNaurois16,S.J.Nolan3,A.Noutsos3,
K.J.Orford3,J.L.Osborne3,M.Ouchrif16,20,M.Panter1,G.Pelletier14,S.Pita6,G.Pu¨hlhofer1,13,M.Punch6,
n
B.C.Raubenheimer15,M.Raue5,J.Raux16,S.M.Rayner3,I.Redondo10,20⋆⋆⋆,A.Reimer17,O.Reimer17,
a
J J.Ripken5,L.Rob18,L.Rolland16,G.Rowell1,V.Sahakian2,L.Sauge´14,S.Schlenker7,R.Schlickeiser17,
3 C.Schuster17,U.Schwanke7,M.Siewert17,H.Sol8,R.Steenkamp19,C.Stegmann7,J.-P.Tavernet16,R.Terrier6,
1 C.G.The´oret6,M.Tluczykont10,20,G.Vasileiadis9,C.Venter15,P.Vincent16,B.Visser15,H.J.Vo¨lk1,and
S.J.Wagner13
1
v
5 1 Max-Planck-Institutfu¨rKernphysik,P.O.Box103980,D69029Heidelberg,Germany
6 2 YerevanPhysicsInstitute,2AlikhanianBrothersSt.,375036Yerevan,Armenia
2 3 UniversityofDurham,DepartmentofPhysics,SouthRoad,DurhamDH13LE,U.K.
1
4 Centre d’Etude Spatialedes Rayonnements, CNRS/UPS,9 av. du Colonel Roche, BP 4346, F-31029 Toulouse Cedex 4,
0
France
5
5 Universita¨tHamburg,Institutfu¨rExperimentalphysik,LuruperChaussee149,D22761Hamburg,Germany
0
/ 6 PhysiqueCorpusculaireetCosmologie,IN2P3/CNRS,Colle`gedeFrance,11PlaceMarcelinBerthelot,F-75231ParisCedex
h 05,France
p 7 Institutfu¨rPhysik,Humboldt-Universita¨tzuBerlin,Newtonstr.15,D12489Berlin,Germany
-
o 8 LUTH,UMR8102duCNRS,ObservatoiredeParis,SectiondeMeudon,F-92195MeudonCedex,France
r 9 Grouped’AstroparticulesdeMontpellier,IN2P3/CNRS,Universite´MontpellierII,CC85,PlaceEuge`neBataillon,F-34095
st MontpellierCedex5,France
a 10 LaboratoireLeprince-Ringuet,IN2P3/CNRS,EcolePolytechnique,F-91128Palaiseau,France
v: 11 DublinInstituteforAdvancedStudies,5MerrionSquare,Dublin2,Ireland
i 12 Serviced’Astrophysique,DAPNIA/DSM/CEA,CESaclay,F-91191Gif-sur-Yvette,France
X 13 Landessternwarte,Ko¨nigstuhl,D69117Heidelberg,Germany
r 14 Laboratoire d’Astrophysique de Grenoble, INSU/CNRS, Universite´ Joseph Fourier, BP 53, F-38041 Grenoble Cedex 9,
a
France
15 UnitforSpacePhysics,North-WestUniversity,Potchefstroom2520,SouthAfrica
16 Laboratoire dePhysiqueNucle´aireet deHautesEnergies, IN2P3/CNRS,Universite´sParisVI &VII,4PlaceJussieu, F-
75231ParisCedex05,France
17 Institut fu¨r Theoretische Physik,Lehrstuhl IV: Weltraumund Astrophysik, Ruhr-Universita¨t Bochum, D 44780 Bochum,
Germany
18 InstituteofParticleandNuclearPhysics,CharlesUniversity,VHolesovickach2,18000Prague8,CzechRepublic
19 UniversityofNamibia,PrivateBag13301,Windhoek,Namibia
20 EuropeanAssociatedLaboratoryforGamma-RayAstronomy,jointlysupportedbyCNRSandMPG
Received?/Accepted?
Abstract.
Veryhighenergy(>100GeV)gamma-rayemissionhasbeendetectedforthefirsttimefromthecompositesupernovaremnant
G0.9+0.1usingtheH.E.S.S. instrument.Thesourceisdetectedwithasignificanceof≈13σ,andaphotonfluxabove200GeV
of (5.7±0.7 ±1.2 ) × 10−12 cm−2s−1,makingitoneoftheweakest sourcesever detectedatTeVenergies. Thephoton
stat sys
spectrum iscompatible withapower law (dN/dE ∝ E−Γ) withphoton index Γ = 2.40±0.11 ±0.20 .Thegamma-ray
stat sys
emission appears to originate in the plerionic core of the remnant, rather than the shell, and can be plausibly explained as
inverseComptonscatteringofrelativisticelectrons.
Keywords.ISM:supernovaremnants–ISM:individualobjects:G0.9+0.1–gamma-rays:observations;
1. Introduction 2. H.E.S.S. ObservationsandResults
The large field of view of the H.E.S.S. cameras (∼5◦) pro-
G0.9+0.1 is a well known composite supernova rem-
videsgoodsensitivityforpointsourcesatanangulardistance
nant, recognised as such from its radio morphology
upto∼2◦ fromthepointingdirectionofthetelescopesystem.
(Helfand & Becker 1987). It exhibits a bright compact core
(∼2′ across) surrounded by an 8′ diameter shell. The radio Thebulkofobservationspresentedhereweretakeninwobble
spectrumofthecoreissignificantlyharder(α≈0.12)thanthat mode around the position of SgrA∗. In this mode 28-minute
of the shell (α ≈−0.77) (LaRosa et al. 2000). X-ray observa- runs are taken pointing 0.5◦ away from the nominal source
inalternatingdirections.Approximately20%ofthedatawere
tions of the nebula with BeppoSAX (Mereghetti et al. 1998),
taken in wobble mode around G0.9+0.1 itself. The observa-
Chandra (Gaensler et al. 2001) and XMM-Newton (Porquet
tions have an average offset of 0.9◦ from the SNR. The off-
et al. 2003) have unambiguously identified the core region
axissensitivityofthesystemderivedfromMonteCarlosimu-
as a pulsar wind nebula (PWN). A plausible candidate for
lationshasbeenconfirmedviaobservationsoftheCrabNebula
the central pulsar is the hard spectrum point source CXOU
J174722.8−280915;howevernopulsedemissionhasbeende- (Aharonianetal.2004c).Thetotallivetimeoftheobservations
afterrunselectionis50hours,fromatotalobservationtimeof
tected. Observationswith XMM have revealed a softening of
60 hours. The mean zenith angle of observation was 18◦, re-
the X-ray spectrum with increasing distance from the centre
sultinginanenergythresholdof170GeVafterstandardcuts.
ofthePWN,asignatureofenergylossofelectronswithinthe
nebula.ThelocationofG0.9+0.1intheGalacticCentre(GC)
region suggests a distance of ∼8.5 kpc (Mezger et al. 1996).
At this distance the size of the shell implies an age of a few
thousandyearsfortheSNR.Atgamma-rayenergiesG0.9+0.1
hasnotbeendetectedpreviously.Itisnotassociatedwithany
EGRETsourceandtheonlypublishedobservationintheTeV
domain is an unconstraining upper limit from the HEGRA
collaboration (flux <4.6 × 10−12 cm−2s−1 above 4.2 TeV,
Aharonianetal.2002).
The High Energy Stereoscopic System (H.E.S.S.) is a
newly completed instrument designed to study astrophysical
gamma radiation in the energy range 100 GeV – 10 TeV.
H.E.S.S.consistsoffourimagingatmosphericCherenkovtele-
scopes(Hinton2004;Bernlo¨hretal.2003;Vincentetal.2003;
Aharonianetal.2004a);showerimagesarerequiredinatleast
twoofthemtotriggerthedetector(Funketal.2004).Usingthe
stereoscopictechniquetheinstrumentreachesanangularreso-
lutionof∼0.1◦andapointsourcesensitivityof1%oftheflux
fromtheCrabNebula(5σin25hours).
Observations of the GC region were made with the par-
tially completeH.E.S.S. system in 2003.These data revealed
emissionfromtheSgrAregion,consistentwiththepositionof
SgrA∗(Aharonianetal.2004b).Withinthesamefieldofview
Fig.1. Gamma-ray point source significance map for the re-
indicationsforasignalfromthesupernovaremnantG0.9+0.1
gionaroundtheGC.ThepositionofG0.9+0.1ismarkedwith
were seen at the 4σ level. A relatively deep H.E.S.S. obser- atriangle.SgrA∗ismarkedwithastar.Thecirclesshowthein-
vation of this region with the complete telescope system was
tegrationregionscontributingtothequotedsignificanceatthe
performed in March–September 2004. The results from this
positionsofthesetwoobjects.Thesixtelescopepointingsare
datasetrelevanttoG0.9+0.1arepresentedhere.Adetaileddis-
shownascrosses.
cussion of the implications of the 2004 dataset for the region
aroundSgrA∗ willbepresentedelsewhere.
Standard event reconstruction was applied to these data:
tail-cuts image cleaning, Hillas parameterisation and stereo-
scopicdirectionreconstructionbasedontheintersectionofim-
Send offprint requests to: J.A. Hinton, e-mail:
ageaxes(Aharonianetal.2004d).Cutsonthescaledwidthand
[email protected]
lengthofimages(optimisedongamma-raysimulationsandoff-
⋆ Universite´ LibredeBruxelles, Faculte´ desSciences, Campusde
sourcedata)aremadetosuppressthehadronicbackground.An
laPlaine,CP230,BoulevardduTriomphe,1050Bruxelles,Belgium
⋆⋆ now at Washington Univ., Department of Physics, 1 Brookings additionalcutontheminimumsizeofimages(200photoelec-
Dr.,CB1105,St.Louis,MO63130,USA trons)ismadetofurthersuppressthebackgroundandselecta
⋆⋆⋆ nowatDepartmentofPhysicsandAstronomy,Univ.ofSheffield, subsetofeventswithasuperiorangularresolution(0.07◦rms),
TheHicksBuilding,HounsfieldRoad,SheffieldS37RH,U.K. butwithanincreasedenergythresholdof350GeV.Eventswith
F.Aharonianetal.:VeryhighenergygammaraysfromthecompositeSNRG0.9+0.1 3
directionsreconstructedwithinanangleθ ofa trialsourcelo-
cation are considered on-source where θ < 0.1◦. A ring of
radius0.5◦ andanarea7timesthatoftheon-sourceregionis
usedtoderiveabackgroundestimate.Theon-sourceandback-
groundregioncounts,togetherwithanormalisationfactorfor
thedifferentacceptanceoftheseregions,areusedtoderivethe
statistical significance of any excess following the likelihood -28.1
methodofLi&Ma(1983).
Figure 1 shows a significance sky map for the field of
view of the H.E.S.S. GC observations derived as described
above. Two sources of very high energy (VHE) gamma-rays
areclearlyvisible:theGalacticCentre(HESSJ1745−290)and
a second source (HESSJ1747−281)coincidentwith the SNR
G0.9+0.1. The statistical significance of HESSJ1745−290is
35σ in this dataset, compared with 11σ in our 2003 data
-28.2
(Aharonianetal.2004b).Thisincreaseisexpected(forasteady
source)fromthegreatersensitivityofthefull4-telescopearray
withrespecttothe2-telescopeconfigurationusedin2003,and
fromtheincreasedexposuretimein2004.
TheH.E.S.S.2003datasetshowedevidenceforgamma-ray
17h47m40s 17h47m20s 17h47m00s
emissionatthepositionofG0.9+0.1atthe4σlevel.Thisevi-
denceisconfirmedbythepresenceofa13σexcessatthesame
positionintheindependent2004dataset.Thegamma-raylike Fig.2. 90 cm radio flux map of G0.9+0.1 from LaRosa
excessiswellfitbythepointspreadfunctionoftheinstrument, et al. (2000) (colour scale), overlayed with contours (at 40%
yieldingabestfitpositionofl=0.872◦ ±0.005◦,b=0.076◦ ± and80%peakbrightness)ofthesmoothedandacceptancecor-
0.005◦, consistentwithinstatistical errorswiththe positionof rected count map of gamma-ray candidates (solid lines). The
the PWN in G0.9+0.1 (l = 0.871◦, b = 0.0772◦) (Gaensler H.E.S.S.dataaresmoothedwithaGaussianofrms0.03◦tore-
etal.2001).Thesystematicpointingerrorisapproximately20′′ duce statistical fluctuations. The simulated point-spreadfunc-
ineachdirectionforthesedata. tion of the instrument is shown with dotted lines (also 40%
The simulated point-spread function of the instrument and 80%). The best fit position of the VHE gamma-ray ex-
provides an adequate description of the excess. Assuming cessisshownwitherrorbarsshowingcombinedstatisticaland
a radially symmetric Gaussian emission region (ρ ∝ systematicerrors.Theinnermostcircle(dashed)illustratesthe
exp(−θ2/2σ2source)) a 95% confidence limit on the extension 95%confidencelimitonthermssizeoftheemissionregion.
σ < 1.3′ is obtained. For emission from a uniform thin
source
shell a limit of 2.2′ on the shell radius can be derived at
the same confidence level. Figure 2 shows a radio map of 3. Discussion
G0.9+0.1 at 90 cm (LaRosa et al. 2000) showing the SNR
shell and the central compact source (colour scale). Contours X-rayemissionfromtheshellofG0.9+0.1isonlymarginally
ofVHEgamma-rayemissionaresuperimposedalongwiththe detected above the local GC diffuse emission with a flux of
point spread function of the instrument for comparison. The 2×10−12ergcm−2s−1intherange2–10keV,andisconsistent
reconstructedgamma-raysourcepositionismarkedasacross. withbothathermalornon-thermalorigin(Porquetetal.2003).
Given the close positional coincidence, we identify the VHE Thenon-thermalfluxfromthePWNis5.8×10−12ergcm−2s−1
gamma-raysourceHESSJ1747−281withG0.9+0.1. inthesameenergyrange,correspondingtoaluminosityof∼5×
Figure 3 shows the reconstructedgamma-rayspectrum of 1034ergs−1 assumingadistanceof8.5kpc.Thelackofstrong
G0.9+0.1 using two sets of event selection cuts. A size cut non-thermalX-ray emission from the shell and the point-like
of80photoelectrons(pe)is appliedtoextendthe spectrumto natureoftheH.E.S.S.detectionargueagainstanoriginofthe
lowerenergies.Thisdatasetcanbefitbyapowerlawinenergy VHEemissionintheshell.ThusthePWNappearstobeamore
with photon index 2.40± 0.11 ± 0.20 and a flux above compellingsourcefortheVHEgamma-rays.
stat sys
200GeVof(5.7±0.7 ±1.2 ) × 10−12cm−2s−1.Thefithas ThetotalpowerradiatedbyG0.9+0.1inVHEgamma-rays
stat sys
aχ2/d.o.f.of3.5/5.Usingtheharderimagesizecutof200pe is∼2×1034ergs−1(at8.5kpc),comparedwith∼4×1034ergs−1
used elsewhere in this paper yields consistent results (photon fortheCrabNebula(at2kpc)intheenergyband0.2–10TeV.
index2.29±0.14 andafluxof(5.5±0.8 )×10−12cm−2s−1 Until now the Crab Nebula represented the only firm identi-
stat stat
above200GeV).Thisfluxrepresentsonly2%ofthefluxfrom fication of VHE emission from a nebula poweredby a young
the Crab Nebula above the 200 GeV spectral analysis thresh- neutronstar.
old.AlthoughG0.9+0.1isoneoftheweakestsourceseverde- A simple one-zone inverse Compton model for the VHE
tectedatTeVenergies,thedetectionpresentedhereisstatisti- gamma-ray emission (Khe´lifi 2002) can be used to explain
callyhighlysignificant(13σ),demonstratingthesensitivityof the observed data. A parent population of accelerated elec-
theH.E.S.S.instrument. tronswithabrokenpower-lawspectrumisassumed.Themag-
4 F.Aharonianetal.:VeryhighenergygammaraysfromthecompositeSNRG0.9+0.1
1)
−
−2 sm −1−2 ) s
m −11
−1 cV10−11 erg c10
Te (Fν
E ( ν
d −12 10−12
N/10 Total
d
Synchrotron
IC on CMBR
IC on IR from dust
10−13 10−13 IC on starlight
10−6 10−4 10−2 1 102 104 106 108 1010 1012 1014
10−14 Energy (eV)
Fig.4. The spectral energy distribution of the PWN in
G0.9+0.1 from radio to VHE gamma-ray. Radio data (trian-
10−1 1 10 gles) are taken from (Helfand & Becker 1987), X-ray data
Energy (TeV) (shaded box) from Porquet et al. 2003. The circles show the
Fig.3. ReconstructedVHEgamma-rayspectrumofG0.9+0.1. H.E.S.S.datafromthiswork.Thesolidcurveshowsafitofa
Emptycirclesshowthespectrumderivedusingtherathertight one-zoneinverseComptonmodeltoalldata.Contributionsof
cutsusedforfigures1and2.Solidsymbolsshowthespectrum the CMBR, IRfromdustand starlightphotonfields to the IC
derivedusinganimagesizecutof80photoelectrons(optimised emissionareshown.
formaximalsignificancefor10%Crabstrengthsources)rather
than200photoelectrons.Thelineshowsapower-lawfittothe starlight ISRF is about 5.7eVcm−3, that is 50% smaller than
80pedataset. the value used in GALPROP. This derived value of B is very
close to the equipartition magnetic field (≈5µG), calculated
fromthewholefittedelectronspectrumandgivenasourceof
netic field strength B is assumed to be uniform within the radius 1′ at 8.5 kpc. Given the observed variation of the X-
PWN and is a free parameter. The maximum energy of the ray spectrumwithin the PWN, this one zone modelis clearly
electron spectrum is kept fixed at 500 TeV as it can not be somewhat simplistic. However, such a model, describing the
derived due to the absence of measurements in the hard X- averagepropertiesofthePWN,candescribetheavailabledata
ray/soft γ-ray band. The synchrotron and inverse Compton with reasonable physicalparameters. Future measurementsin
(IC) emission of these electrons are calculated according to theGeVregimewithGLAST(sensitivity2×10−12ergs−1cm−2
Blumenthal& Gould (1970). The Klein-Nishinaeffect on the at 1 GeV) are needed to resolve the ambiguity between the
ICcrosssectionistakenintoaccount. magnetic energy density and that of target radiation fields in
For sources in the GC region the presence of radi- this region. Alternative explanationsbased on a hadronic ori-
ation densities considerably higher than typical Galactic ginoftheVHEemission(forexampleviathedecayofneutral
Plane values leads to substantially enhanced IC emission pionsproducedin p–pcollisions)arenotexcluded.
(deJageretal.1995).TheseedphotonsforICscatteringarethe ThedetectionofVHEgamma-rayemissionfromG0.9+0.1
cosmic microwave background radiation (CMBR), the galac- provides the first direct evidence for the acceleration of very
ticdiffuseemissionaround100µmfromdustandaround1µm energetic particles in this object. The position and point-like
fromstarlight.Theenergyspectraofthedustandstarlightcom- nature of the gamma-ray emission, combined with the broad
ponents are taken from a recent estimation of the interstellar bandSEDofG0.9+0.1,stronglysuggestanoriginoftheVHE
radiation field (ISRF) used for the GALPROP code (Strong emission in the compact central source (previously identified
et al. 2000). To account for possible local variations of the asapulsarwindnebula).
starlight component the energy density of this component is This detection represents the first step towards the study
keptasafreeparameter.TheenergydensityofthedustISRFis ofanemergingpopulationofVHEgamma-rayemittingPWN.
keptfixedat0.23eVcm−3 asusedinGALPROP.Thestarlight ThedetectionofnewPWNwithaccurateTeVspectraandwith
ISRF energydensity, the electron spectrum parametersand B morphologicalmeasurementswill provide informationkey to
areadjustedsuchthatthecomputedsynchrotronandICspectra understandingparticleaccelerationinthevicinityofpulsars.
matchtheobservations.
Figure4showsthebroad-bandspectralenergydistribution
Acknowledgements
(SED) of G0.9+0.1,together with the best modelfit. The fit-
tedmeanmagneticfieldstrengthis≈6µG.Themodelelectron ThesupportoftheNamibianauthoritiesandoftheUniversity
spectrumhasalowenergyindexof0.6andahighenergyindex of Namibia in facilitating the construction and operation of
of2.9withabreakat25GeV.Thebestfitenergydensityofthe H.E.S.S. is gratefully acknowledged, as is the support by
F.Aharonianetal.:VeryhighenergygammaraysfromthecompositeSNRG0.9+0.1 5
the German Ministry for Education and Research (BMBF),
the Max Planck Society, the French Ministry for Research,
the CNRS-IN2P3 and the Astroparticle Interdisciplinary
Programme of the CNRS, the U.K. Particle Physics and
Astronomy Research Council (PPARC), the IPNP of the
Charles University, the South African Department of Science
and Technology and National Research Foundation, and by
the University of Namibia. We appreciate the excellent work
of the technical support staff in Berlin, Durham, Hamburg,
Heidelberg,Palaiseau,Paris,Saclay,andinNamibiainthecon-
struction and operationof the equipment.We would also like
tothankJ.Lazio(NRL)forprovisionof90cmradiodataand
A. W. Strong for providing a parameterisation of interstellar
radiationfields.
References
Aharonian, F., Akhperjanian, A. G., Beilicke, M., 2002, Astron. &
Astrophys.,395,803.
Aharonian, F., Akhperjanian, A. G., Aye, K.-M., et al., 2004a,
Astropart.Phys.,22,109.
Aharonian,F.,Akhperjanian,A.G.,Aye,K.-M.,etal.,2004b,Astron.
&Astrophys.,425,L13.
Aharonian, F., Akhperjanian, A. G., Aye, K.-M., et al., 2004c, In
preparation.
Aharonian, F., Akhperjanian, A. G., Aye, K.-M., et al., 2004d,
AcceptedbyAstron.&Astrophys.(astro-ph/0411582)
Bernlo¨hr,K.,Carrol,O.,Cornils,R.,etal.,2003,APh.,20,111.
Blumenthal,G.R.,Gould,R.J.,1970,RvMP.,42,Number2,237.
Funk, S., Hermann, G., Hinton, J., et al, 2004, Astropart. Phys. 22,
285.
Gaensler,B.M.,Pivovaroff,M.J.,Garmire,G.P.,2001,Astrophys.J.,
556,L107.
Helfand,D.J.,Becker,R.H.,1987,Astrophys.J.,314,203.
Hinton, J.A. (H.E.S.S. Collaboration), 2004, New Astron. Rev., 48,
331.
de Jager, O.C., Harding, A.K., Baring, M.G. and Mastichiadis, A.,
1995,Proc.24thICRC(Rome)p528.
Khe´lifi, B., 2002, PhD Thesis, University of Caen, France
(http://lpnp90.in2p3.fr∼cat/Thesis/)
LaRosaT.N.,KassimN.E.,LazioT.J.W.,HymanS.D.,2000,Astron.
J.,119,207.
Li,T.&Ma,Y.,1983,Astrophys.J.,272,317.
Mereghetti, S., Sidoli, L., Israel, G.L., 1998, Astron. & Astrophys.,
331,L77.
Mezger,P.G.,Duschl,W.J.,Zylka,R.,1996A&ARv.,7,289.
Porquet, D., Decourchelle, A., Warwick, R.S., 2003, Astron. &
Astrophys.,401,197.
Strong,A.W.,Moskalenko,I.V.,Reimer,O.,2000,ApJ,537,763S.
VincentP.,DenanceJ.-P.,HuppertJ.-F.,etal.2003,Proc.28thICRC,
Tsukuba,Univ.AcademyPress,Tokyo,p.2887.