Detailed spectral and morphological analysis of the shell type SNR RCW 86

0.67 MB·

by  H. E. S. S. Collaboration

Astronomy&Astrophysicsmanuscriptno.RCW86˙archive (cid:13)c ESO2016
January19,2016
Detailed spectral and morphological analysis of the shell type SNR
RCW 86
H.E.S.S.Collaboration,A.Abramowski1,F.Aharonian2,3,4,F.AitBenkhali2,A.G.Akhperjanian5,4,E.O.Angu¨ner6,
M.Backes7,A.Balzer8,Y.Becherini9,J.BeckerTjus10,D.Berge11,S.Bernhard12,K.Bernlo¨hr2,E.Birsin6,
R.Blackwell13,M.Bo¨ttcher14,C.Boisson15,J.Bolmont16,P.Bordas2,J.Bregeon17,F.Brun18,P.Brun18,M.Bryan8,
T.Bulik19,J.Carr20,S.Casanova21,2,N.Chakraborty2,R.Chalme-Calvet16,R.C.G.Chaves17,22,A,Chen23,
J.Chevalier24,M.Chre´tien16,S.Colafrancesco23,G.Cologna25,B.Condon26,J.Conrad27,28,C.Couturier16,Y.Cui29,
I.D.Davids14,7,B.Degrange30,C.Deil2,P.deWilt13,A.Djannati-Ata¨ı31,W.Domainko2,A.Donath2,L.O’C.Drury3,
G.Dubus32,K.Dutson33,J.Dyks34,M.Dyrda21,T.Edwards2,K.Egberts35,P.Eger2,J.-P.Ernenwein20,P.Espigat31,
6 C.Farnier27,S.Fegan30,F.Feinstein17,M.V.Fernandes1,D.Fernandez17,A.Fiasson24,G.Fontaine30,A.Fo¨rster2,
1 M.Fu¨ßling36,S.Gabici31,M.Gajdus6,Y.A.Gallant17,T.Garrigoux16,G.Giavitto36,B.Giebels30,J.F.Glicenstein18,
0 D.Gottschall29,A.Goyal37,M.-H.Grondin26,M.Grudzin´ska19,D.Hadasch12,S.Ha¨ffner38,J.Hahn2,J.Hawkes13,
2
G.Heinzelmann1,G.Henri32,G.Hermann2,O.Hervet15,A.Hillert2,J.A.Hinton2,33,W.Hofmann2,P.Hofverberg2,

n C.Hoischen35,M.Holler30,D.Horns1,A.Ivascenko14,A.Jacholkowska16,M.Jamrozy37,M.Janiak34,
a
J F.Jankowsky25,I.Jung-Richardt38,M.A.Kastendieck1,K.Katarzyn´ski39,U.Katz38,D.Kerszberg16,B.Khe´lifi31,
8 M.Kieffer16,S.Klepser36,D.Klochkov29,W.Kluz´niak34,D.Kolitzus12,Nu.Komin23,K.Kosack18,S.Krakau10,
1 F.Krayzel24,P.P.Kru¨ger14,H.Laffon26,G.Lamanna24,J.Lau13,J.Lefaucheur31,V.Lefranc18,A.Lemie`re31,
M.Lemoine-Goumard26,J.-P.Lenain16,T.Lohse6,A.Lopatin38,M.Lorentz18,C.-C.Lu2,R.Lui2,V.Marandon2,
]
E A.Marcowith17,C.Mariaud30,R.Marx2,G.Maurin24,N.Maxted17,M.Mayer6,P.J.Meintjes40,U.Menzler10,
H M.Meyer27,A.M.W.Mitchell2,R.Moderski34,M.Mohamed25,K.Morå27,E.Moulin18,T.Murach6,M.deNaurois30,
. J.Niemiec21,L.Oakes6,H.Odaka2,S.O¨ttl12,S.Ohm36,B.Opitz1,M.Ostrowski37,I.Oya36,M.Panter2,
h
R.D.Parsons2,M.PazArribas6,N.W.Pekeur14,G.Pelletier32,P.-O.Petrucci32,B.Peyaud18,S.Pita31,H.Poon2,
p
- D.Prokhorov9,H.Prokoph9,G.Pu¨hlhofer29,M.Punch31,A.Quirrenbach25,S.Raab38,I.Reichardt31,A.Reimer12,
o O.Reimer12,M.Renaud17,R.delosReyes2,F.Rieger2,41,C.Romoli3,S.Rosier-Lees24,G.Rowell13,B.Rudak34,
r
t C.B.Rulten15,V.Sahakian5,4,D.Salek42,D.A.Sanchez24,A.Santangelo29,M.Sasaki29,R.Schlickeiser10,
s
a F.Schu¨ssler18,A.Schulz36,U.Schwanke6,S.Schwemmer25,A.S.Seyffert14,R.Simoni8,H.Sol15,F.Spanier14,
[ G.Spengler27,F.Spies1,Ł.Stawarz37,R.Steenkamp7,C.Stegmann35,36,F.Stinzing38,K.Stycz36,I.Sushch14,
1 J.-P.Tavernet16,T.Tavernier31,A.M.Taylor3,R.Terrier31,M.Tluczykont1,C.Trichard24,R.Tuffs2,K.Valerius38,
v J.vanderWalt14,C.vanEldik38,B.vanSoelen40,G.Vasileiadis17,J.Veh38,C.Venter14,A.Viana2,P.Vincent16,
1 J.Vink8,F.Voisin13,H.J.Vo¨lk2,T.Vuillaume32,S.J.Wagner25,P.Wagner6,R.M.Wagner27,M.Weidinger10,
6
R.White33,2,A.Wierzcholska25,21,P.Willmann38,A.Wo¨rnlein38,D.Wouters18,R.Yang2,V.Zabalza33,D.Zaborov30,
4
4 M.Zacharias25,A.A.Zdziarski34,A.Zech15,F.Zefi30,andN.Z˙ywucka37
0
(Affiliationscanbefoundafterthereferences)
.
1
0 AcceptedNovember11,2015
6
1
: ABSTRACT
v
i
X Aims. We aim for an understanding of the morphological and spectral properties of the supernova remnant RCW 86 and for insights into the
productionmechanismleadingtotheRCW86veryhigh-energyγ-rayemission.
r
a Methods.WeanalyzedHighEnergySpectroscopicSystem(H.E.S.S.)datathathadincreasedsensitivitycomparedtotheobservationspresented
in the RCW 86 H.E.S.S. discovery publication. Studies of the morphological correlation between the 0.5 – 1 keV X-ray band, the 2 – 5 keV
X-rayband,radio,andγ-rayemissionshavebeenperformedaswellasbroadbandmodelingofthespectralenergydistributionwithtwodifferent
emissionmodels.
Results. We present the first conclusive evidence that the TeV γ-ray emission region is shell-like based on our morphological studies. The
comparisonwith2–5keVX-raydatarevealsacorrelationwiththe0.4–50TeVγ-rayemission.ThespectrumofRCW86isbestdescribed
byapowerlawwithanexponentialcutoffatE = (3.5±1.2 )TeVandaspectralindexofΓ ≈ 1.6±0.2.Astaticleptonicone-zonemodel
cut stat
adequatelydescribesthemeasuredspectralenergydistributionofRCW86,withtheresultanttotalkineticenergyoftheelectronsabove1GeV
beingequivalentto∼0.1%oftheinitialkineticenergyofaTypeIasupernovaexplosion(1051 erg).Whenusingahadronicmodel,amagnetic
field of B ≈ 100 µG is needed to represent the measured data. Although this is comparable to formerly published estimates, a standard E−2
spectrumfortheprotondistributioncannotdescribetheγ-raydata.Instead,aspectralindexofΓ ≈ 1.7wouldberequired,whichimpliesthat
p
∼ 7×1049/ncm−3erghasbeentransferredintohigh-energyprotonswiththeeffectivedensityncm−3 = n/1cm−3.Thisisabout10%ofthekinetic
energyofatypicalTypeIasupernovaundertheassumptionofadensityof1cm−3.
Keywords.γ-rays:observations;individual(RCW86,G315.4-2.3);supernovaremnants
1
1. Introduction
Supernova remnants (SNR) are prime candidates to be sources
ofGalacticcosmicrays.Thedetectionofvery-high-energyγ-ray
(VHE;E>100GeV)andnonthermalX-rayemissionfromSNRs
has shown that they do in fact accelerate particles to energies
above100TeV(see,e.g.,Hinton&Hofmann2009;Aharonian
2013).
RCW 86 (also known as G315.4-2.3 or MSH 14-63) is lo-
cated at a distance of (2.5 ± 0.5) kpc (Helder et al. 2013). It
isalmostcircularinshapewithadiameterofabout40’clearly
showingashell-likestructureintheoptical(Smith1997),radio
(Kesteven&Caswell1987),infrared(Williamsetal.2011)and
X-ray (Pisarski et al. 1984) regimes. The source has been de-
tectedinγ-rays(Aharonianetal.2009;Yuanetal.2014),buta
shell-likestructurewasnotresolved.RCW86isayoungSNR,
which is about 1800 years old, based on the probable connec-
tiontothehistoricalsupernovaSN185recordedbyChineseas-
tronomersin185AD(Stephenson&Green2002;Smith1997;
Dickeletal.2001;Vinketal.2006).ThenatureoftheRCW86
supernova (SN) progenitor was under intense discussion in the
context of its young age and its unusually large size with a ra-
dius of R ≈ 15d2.5 pc (d2.5: the distance to RCW 86 in units Fig.1. VHE γ−ray emission of RCW 86. The sky map is ex-
of 2.5 kpc). Williams et al. (2011) comprehensively studied all tracted with faint analysis cuts and smoothed with a Gaussian
argumentsaboutthetypeoftheprogenitorofRCW86andcon- filterwithσ =0◦.06toreducetheeffectofstatisticalfluctu-
smooth
ductedhydrodynamicsimulationstoexplaintheobservedchar- ations.Blackcontourscorrespondto3,5,7σsignificance.The
acteristics. They concluded that RCW 86 is likely the remnant whitedottedcircledepictstheintegrationregionchosenforthe
of a Type Ia supernova. The explosion was off-center in a low- spectralanalysis.Thedashedcyansectorshowsthepositionan-
density cavity carved by the progenitor system (see also Vink gle range of the radial profile in Fig. 3. The two green sectors
etal.1997;Broersenetal.2014). (solid lines) give the extraction regions for spectra of the SW
ThephysicalconditionsvarywithinthevolumeofRCW86. andNEregionsdiscussedinSect.3.
While slow shocks have been measured (∼600 – 800kms−1;
Long & Blair 1990; Ghavamian et al. 2001) in the southwest
(SW) and northwest (NW) regions along with relatively high
post-shock gas densities (∼2 cm−3; Williams et al. 2011), X-
ray measurements by Helder et al. (2009) indicate high veloc-
ities of 6000 ± 2800kms−1 in the northeast (NE) whereas op- Initsfirstphase,thesystemconsistedoffouridenticalimaging
tical measurements of specific knots in this region by Helder Cherenkovtelescopes,eachwithamirrorareaof107m2 anda
et al. (2013) show large variations in proper motions between large field of view of 5◦ (Bernlo¨hr et al. 2003). The data pre-
700 – 2000kms−1 and low densities of ∼0.1 – 0.3 cm−3 (see sentedinthispaperweretakenduringthisphase.
Yamaguchietal.2011).IntheSW,NE,andtoalesserextentin
In 2009 the H.E.S.S. Collaboration announced the discov-
theNWregion,nonthermalX-rayemissionisdetected(Bamba
eryofRCW86intheVHEregime(Aharonianetal.2009).The
etal.2000;Vinketal.1997,2006;Williamsetal.2011).Near
sourcemorphologyofRCW86wasthoroughlystudied,butthe
a region of Fe-K-line emission in the SW, there is nonthermal
existingdatadidnotsufficetosettlethequestionofwhetherthe
emissionpossiblysynchrotronradiationfromelectronsacceler-
VHEemissionoriginatesfromashellorfromasphericalregion.
atedatthereverseshock(Rhoetal.2002).
Between2007and2011,observationsoftheneighboringsource
Here, we present a new analysis of very high-energy γ-
HESS J1458−608 (de Los Reyes Lopez 2011) and the scan of
ray data of RCW 86 taken with the High Energy Stereoscopic
theGalacticplanehaveaddedadditionalobservationtimetothe
System (H.E.S.S.). When the discovery of RCW 86 was pub-
data set, which now amounts to ∼ 57 h. The zenith angles of
lishedbytheH.E.S.S.collaboration,themorphologyintheTeV
theseobservationsrangefrom36◦to53◦andhaveameanvalue
γ−ray regime could not be resolved owing to limited statistics
of 40◦. The data were recorded with pointing offsets between
(Aharonian et al. 2009). The new analysis benefits from sig-
0.5◦and2.5◦,average1.1◦,fromtheRCW86position.Thesen-
nificantly improved statistics and addresses whether RCW 86
sitivity and angular resolution were improved compared to the
has a shell structure in the VHE γ-ray regime. In addition, the
former publication (Aharonian et al. 2009), using an advanced
larger data set allows for a broadband modeling of the spectral
analysis method (for further detail, see de Naurois & Rolland
energy distribution with both a leptonic and an hadronic one-
2009) instead of the formerly used standard Hillas-parameter-
zonemodel.Werestrictedourselvestotwosimplemodelssince
basedanalysis.Standardcutswereusedforspectralanalysisand
it is unreasonable to use more intricate models with additional
faint source cuts for morphological analysis. The faint source
parametersthatcannotbedeterminedowingtolimitedstatistics.
cutshaveanimprovedangularresolutionattheexpenseoflower
statisticsandahigherenergythresholdofabout100GeV(forthe
cut definition, see de Naurois & Rolland 2009). The results we
2. H.E.S.S.observationsandanalysismethod
presentareconsistentwiththeresultsofamultivariateanalysis
TheHighEnergyStereoscopicSystemislocatedintheKhomas technique (Ohm et al. 2009), using an independent calibration
HighlandofNamibiaataheightofabout1800mabovesealevel. scheme.
H.E.S.S.Collaborationetal.:DetailedanalysisoftheshelltypeSNRRCW86
10−12cm−2s−1TeV−1 (Log(likelihood) = −54; see Fig. 2). We
True energy (TeV)
1 10 102 tested if a power law with an exponential cutoff dN/dE =
-1)V Φ0(E/1TeV)−Γexp(−E/Ecut) describes the data better than a
e10-11 simplepowerlaw.ThefitgaveΦ(1TeV)=(3.0±0.2 ±0.6 )×
T stat sys
-2-1 sm10-12 1er0g−y12Ecmcut−=2s−(13T.5eV±−11.2wstiatth±Γ2=.21sy.6s)±Te0V.2s(tLato±g0(.l2iksyesliahnodoad)cu=to−ff4e7n.7-
ux (c10-13 )o,fw3h.5icσh,iascfcaovrodriendgotvoeartlhoeg-sliimkeplliehopoodwreartiloawtewst.itThhaessigpnecifitrcaalnrcee-
Fl10-14 sultofthepower-lawfitiscompatiblewithapreviousH.E.S.S.
publication(Aharonianetal.2009).
10-15
10-16 3.2. Morphology
To settle the question of whether the TeV γ-ray emission is
10-17
shell-likeasfoundintheoptical,radio,andX-raybands,atwo-
dimensional log-likelihood fit was applied to the uncorrelated
10-18
sky maps taking the morphological model and instrument re-
sponse into account. To distinguish between different morpho-
10-19
logicalstructures,weusedtwoalternativemodels:asphereand
)10-204 ashellmodel.Themodelsareobtainedbyline-of-sightintegrals
c 3 of a three-dimensional emitting sphere or shell. In the case of
Ds ( 2 theshellmodel,thefreeparametersarethecenterpositionand
al theinnerandouterradius.Thespheremodelisderivedfromthe
u 1
d shellmodelbyfixingtheinnerradiustozero.Bothmodelshave
si 0
e uniformemissivity.Table1summarizesthefitresults.Theshell
R -1
structureisfavoredoverthespherestructurewithasignificance
-2
ofabout8σ.InFig.3(leftpanel),theradialprofileoftheTeV
-3
emissionitselfisgiventogetherwiththefitresultsofbothmod-
-4
1 10 102 els.
Reconstructed energy (TeV)
To facilitate the comparison, radial excess profiles of VHE
Fig.2. In the upper panel, the differential energy spectrum of γ-ray,X-ray,andradioemissionarepresentedinthecenterand
RCW86withthebest-fitexponentialcutoffpower-lawmodelis right panels of Fig. 3. The X-ray data is split into two energy
bands: 0.5 – 1 keV and 2 – 5 keV. The latter exhibits a higher
shown. The error bars denote 1σ statistical errors. The shaded
amountofnonthermalemissionincomparisontothelow-energy
arearepresentsthe1σconfidencelevelofthefittedspectrum.In
band. The nonthermal emission in RCW 86 is not evenly dis-
thebottompanel,thecorrespondingresidualsaregiven.
tributed. Broersen et al. (2014) have shown that the NE region
isdominatedbynonthermalemission,whiletheSWregionex-
hibitssimilaramountsofhotthermalandnonthermalemission.
3. Results IntheNWandSEregions,however,thermalemissionprevails.
InFig.3,thecentralpaneldepictstheradialprofileofthewhole
The γ-ray excess map, using the ring background subtraction
SNR, whereas in the right panel only the data of the SW re-
technique(Bergeetal.2007),isshowninFig.1.Themapwas
smoothed with a Gaussian filter to reduce the effect of statis- gionwithinthepositionanglerange1 of190◦ to230◦ wereused
tical fluctuations. The 68% containment radius of 0◦.061 of the to determine the profiles. The radio data are from Molonglo
Observatory Synthesis Telescope (MOST; Whiteoak & Green
H.E.S.S. point spread function (PSF) was chosen as smoothing
1996) and the X-ray data are taken with the XMM–Newton X-
radius.
ray telescope (Broersen et al. 2014). The radial profile of the
Sincethisislargerthanthebinsize(0◦.02),thepixelsarecor-
wholeremnant(Fig.3,centerpanel)clearlyshowsthattheTeV
relatedwitheachother.Thisisaccountedforinthespatial-fitting
andtheX-rayemissionoftheenergyrangebetween2–5keV
process that we present. Extended γ-ray emission was found.
arecorrelated,whereasinthecaseoflowenergeticX-rayemis-
We detected 1220 ± 87 excess γ-rays within a circular region
sion (0.5 – 1 keV) and radio emission a weaker luminosity is
centered at α = 14h43m2.16s, δ = −62◦26(cid:48)56(cid:48)(cid:48) with a
J2000 J2000 found in the central region and the shell-like emission is more
radius of 0◦.41 (see Fig. 1) and a source detection at a statisti-
pronounced.TheradialprofilesoftheSWregionofTeVγ-ray
calsignificanceof18.3σ.Thecenterisgivenbythefluxaver-
emission, low- and high-energy X-ray radiation (Fig. 3, right
agedcentroid,whichwasdeterminedwiththespectralresultsof
panel)againseemtorevealacorrelationbetweenthe2–5keV
thepreviousH.E.S.S.publicationofRCW86(Aharonianetal.
X-rayemissionandtheTeV-emission.Inaddition,itseemsthat
2009).Theradiuswasadjustedtotheimprovedanalysismethod
themaximumofthe2–5keVX-rayemissionisslightlymore
andbetterangularresolution.
off-center than the maximum of the TeV emission, but this is
notsignificantbecauseofthelowstatisticsoftheTeVdata.The
3.1. Spectralanalysis maximaofthelow-energyX-ray(0.5–1keV)andradioemis-
sion are at larger radii. In summary, the TeV emission and the
A fit of a power-law function dN/dE = Φ0(E/1TeV)−Γ to high-energyX-rayemission(2–5keV)arecorrelatedthrough-
the spectral data analyzed with the standard analysis cuts outtheremnant,whereasradioandlow-energyX-rayemission
yields Γ = 2.3 ± 0.1 ± 0.2 , and the differential flux
stat sys
normalization at 1 TeV of Φ0 = (2.9 ± 0.2stat ± 0.6sys) × 1 Positionangle0◦correspondstonorthand90◦toeast
3
H.E.S.S.Collaborationetal.:DetailedanalysisoftheshelltypeSNRRCW86
excess/area (a.u)00..0034..5545 0gSS(cid:176)-ph r<ahe yleFl rdme <a om t3ado6e0dl(cid:176)e l excess/area (a.u)000...345 0gRXX-(cid:176)--ar rr<adaa yiyyoF d20 a<-.55t a3- k16e 0kV(cid:176)e V excess/area (a.u)00..45 1gRXX-9--arrr0adaa(cid:176)yiyy o< d20 aF-.55t a- <k1 e 2kV3e0V(cid:176)
0.3 0.3
0.25
0.2
0.2 0.2
0.15
0.1 0.1
0.1
0.05 0 0
00 0.1 0.2 0.3 0.4 0.5 0.6 0 0.1 0.2 0.3 0.4 0.5 0.6 0 0.1 0.2 0.3 0.4 0.5 0.6
r ((cid:176)) r ((cid:176)) r ((cid:176))
Fig.3.Radialprofiles:Theshellmodel’sbest-fitcenterpointposition(seeTabel1)servesasacommoncenterfortheradialprofiles
of the γ-ray, X-ray, and radio data. In left panel: The radial profile of the spatial distribution of the TeV γ-ray emission is shown
asblackcrosses,whichdepictthemeasuredTeVdatapointsandtheirerrors.Thedashedredlinecorrespondstotheradialprofile
ofthespheremorphologyandthegreenlinetotheshellmodel.Incenterandrightpanel:RadialprofilesintheTeVγ-ray,radio,
0.5 – 1 keV, and 2 – 5 keV X-ray regimes for different regions of RCW 86 are shown. While the center panel covers the entire
position angle range, the right panel only shows the SW region with position angles between 190◦ and 230◦ (position angle 0◦
corresponds to north and 90◦ to east; see Fig. 1). Black crosses are measured VHE excess points. The low-energy (0.5 – 1 keV,
dashed cyan line) and high-energy ( 2 – 5 keV, dotted brown line) X-ray band data and the radio data (solid magenta line) were
smoothedwiththeH.E.S.S.-PSFtoaccountforthedifferentangularresolutionofthedifferentinstruments.Theradiodataarefrom
MolongloObservatorySynthesisTelescope(MOST;seeWhiteoak&Green1996)andtheX-raydataarefromtheXMM-Newton
X-raytelescope(Broersenetal.2014).Theγ-rayandX-raydataarenormalizedsothattheareaunderneaththecurveisequalto
one;theradiodataarescaledsuchthattheareaunderneaththedatapointsisequalto0.5.
Table1.Resultsofthetwo-dimensionallog-likelihoodfitofmorphologicalmodelstotheH.E.S.S.data.(α ,δ )
center,J2000 center,J2000
givesthecenterofthesphereandshellmodels.Theparametersr andr standfortheinnerandouterradiusofthesphere.Inthe
in out
caseofthesphere,r iskeptatzeropc.
in
Model α δ r [pc] r [pc] Log(likelihood)
center,J2000 center,J2000 in out
sphere 14h42m58s±0h0m9s −62◦25(cid:48)48(cid:48)(cid:48)±0◦1(cid:48)1(cid:48)(cid:48) – 15.9±0.6 −4093.35
shell 14h42m53s±0h0m7s −62◦25(cid:48)48(cid:48)(cid:48)±0◦1(cid:48)48(cid:48)(cid:48) 9.5±1.4 14.8±1.0 −4046.82
regionsarefurtherawayfromthecenterandshowlowerlevels 4. Discussion
ofemissioninthecentralregionoftheremnant.
SNRs are thought to be the primary sources for the bulk of
Galacticcosmic-rayprotonswithenergiesupto∼3PeV,butthe
finalproofisstilllacking.Thebroadband,nonthermalemission
of these sources is produced by accelerated particles through
To study the influence of the different physical conditions
several channels, e.g., synchrotron, inverse Compton, nonther-
like shock speeds and densities on the γ-ray emission, spec-
malBremsstrahlung,andneutralpiondecay.Toachieveinsights
tra of two different subregions (SW and NE) were extracted. intotheγ-rayproductiontakingplaceinSNRs,thedifferentpro-
Figure 1 shows the regions, their position, and their size. They
cesses have to be disentangled and analyzed. We present the
are both centered around the best-fit position of the 3D-shell results of the broadband data of RCW 86 in two different sce-
model. A Gaussian function was fitted to the radial profile of
narios for radiation mechanism for the γ-ray emission: inverse
the γ-ray data and the 1σ interval around the mean of the fit Compton (IC) scattering of electrons off ambient photons (lep-
was then taken as the minimal and maximal radii of the re-
tonic scenario) or proton-proton interaction with the ambient
gions. In both regions, a detection significance of about 10 σ
medium (hadronic scenario). For both cases, we have modeled
has been achieved. The resulting spectrum is well described
the broadband emission of RCW 86 with a simple static one-
by a power-law model, which was fitted to the data up to an
zone model (presented in Acero et al. 2010), where the radi-
energy of 20 TeV. It is not possible to distinguish between a ation of the different wavebands is produced within the same
simple power-law spectrum and one with an exponential cut-
regionwithaconstantmagneticfield.Thecurrentenergydistri-
off because of the lower statistics of the quadrant data. We ob- butionsoftheparticles(electronsand/orprotons)aregivenbya
tained spectral indices of Γ = (2.5 ± 0.2stat ± 0.2sys) and of powerlawwithanexponentialcutoff.Synchrotronradiationand
Γ=(2.2±0.1 ±0.2 )aswellasadifferentialfluxnormaliza-
stat sys IC scattering on the cosmic microwave background and the lo-
tionat1TeVofΦ0 =(0.7±0.1stat±0.1sys)×10−12cm−2s−1TeV−1 calinterstellar(opticalandinfrared)radiationfields(seePorter
andΦ =(0.7±0.1 ±0.1 )×10−12cm−2s−1TeV−1intheNE &Strong2005)weretakenintoaccountwithenergydensitiesof
0 stat sys
andSWregions,respectively.Thus,thereisnosignificantspec- 0.66eVcm−3(dust)and0.94eVcm−3(stars).Inthecaseofthe
tral difference between both regions. The results are consistent hadronicscenario,theγ-rayproductionwascalculatedfollowing
withthespectrumofthewholeremnant. Kelner et al. (2006). Nonthermal Bremsstrahlung is neglected
4
H.E.S.S.Collaborationetal.:DetailedanalysisoftheshelltypeSNRRCW86
becauseofthelowambientdensity≤1cm−3.Themodelingwas the expectation that electrons and protons exhibit the same dy-
doneundertheassumptionthatthedistanceofRCW86is2500 namics at relativistic energies up to an energy where electron
pc,theouterradiusis15pc,andtheshellthicknessis5pc.The synchrotronlossesbecomeimportant.
emissionatallwavelengthswascalculatedfortheshellregion. Lemoine-Goumard et al. (2012) applied a two-zone model
Thestationaryone-zonemodelishoweveranoversimplification to the data to overcome this problem. They introduced a sep-
because the radio morphology and the nonthermal X-ray mor- arate second leptonic population to explain the radio emission
phologydiffer,whichcannotbeexplainedbythiskindofmodel. withtheconsequencethatthelowerlimitonthemagneticfield
ThespectralenergydistributionpresentedinFig.4iscomposed was 50 µG. The X-ray emission was reproduced with an in-
of the VHE γ-ray spectra presented in Sect. 3, the Fermi-LAT jection spectrum with a power-law index of Γ = 1.8, a break
data points (Yuan et al. 2014), the X-ray spectra of ASCA and at 3 TeV, caused by synchrotron cooling, and an exponen-
RXTE(Lemoine-Goumardetal.2012),andtheradiodatafrom tial cut-off at 20 TeV. The energy injected into hadrons was
the Molonglo at 408 MHz and Parkes at 5 GHz (Caswell et al. ∼7×1049/n erg.Theseresultsfulfilltheobservationalcon-
cm−3
1975;Lemoine-Goumardetal.2012). straints,buttheproblemremainsthattheprotonspectrumispar-
Intheleptoniccase,thebroadbanddatacanbedescribedby ticularlyhard.
aone-zonemodelusinganelectronspectrumwithaspectralin- Whenwecomparethemodelingresultsofboththeleptonic
dex of Γ ≈ 2.3, an exponential cutoff at E ≈ 19 TeV, and a andhadronicmodels,wefindthatneithercanberuledout.The
e cut
magneticfieldofB≈22µG(seeFig.4).Theseresultsarecom- hadronicmodelencountersproblemsbydescribingthedataina
parable to those of Lemoine-Goumard et al. (2012) and Yuan self-consistent way, i.e., it is not possible to describe the radio
et al. (2014) and the magnetic field is comparable to estimates datawithelectronandprotonspectrawiththesamespectralin-
by Vink et al. (2006). The total kinetic energy of all electrons dex,whichwouldbeexpectedforrelativisticparticlesunderthe
above1GeVamountsto∼1048 erg,whichisabout0.1%ofthe assumptionofdiffuseshockaccelerationtheory.Betterstatistics
totalenergyofatypicalTypeIasupernova(1051erg).Weobtain andpossiblymoredetailedmodelsareneededtosolvethisques-
anupperlimitfortheenergyinjectedintoacceleratedprotonsof tion.
∼1049/n ergwiththeeffectivedensityn =n/1cm−3also Our study of the morphological correlation between radio,
cm−3 cm−3
takingintoaccountprotonswiththesamespectralindexasthe X-rays,andγ-rayshasshown(seeSect.3)thatradioandtheX-
electronsandaconservativechosenenergycutoffof100TeV.A rayemissionintheenergyrangebetween0.5and1keVdonot
higheramountofenergyisnotcompatiblewiththeFermi-LAT coincidewiththeγ-rayemission,whiletheemissionintheX-ray
upperlimits(seeFig.4).Thisenergylimitimpliesanelectron- regime2–5keVshowsacorrelationwiththeTeVγ-rayemis-
to-protonratioabove1GeVofK ≥ 10−1.Theseresultsarein sion.ThisX-rayemission(2–5keV)isspatiallyneartoaFe-
ep
goodagreementwiththoseobtainedbyLemoine-Goumardetal. K-line.Rhoetal.(2002)arguedthatthehardX-raycontinuum
(2012). issynchrotronradiationproducedbyelectrons,whichareaccel-
Inthecaseofthehadronicscenario,theγ-raysareproduced eratedinthereverseshock.Itisthereforepossiblethatsomeof
via proton-proton interactions and subsequent neutral pion de- the VHE gamma-ray emission comes from the region shocked
caywhereastheX-raysarestillproducedviasynchrotronradi- bythereverseshock.Apossiblehintforthisisprovidedbythe
ation of high-energetic electrons. Therefore, the electron frac- morecentrallyfilledmorphologyoftheVHEgamma-rayemis-
tion has to be lower and the magnetic field stronger. A stan- sionwithrespecttotheradioemission,however,thesensitivity
dard proton spectrum with Γ = 2 and a total energy in accel- ofthegamma-rayisnotsufficienttodrawanyfirmconclusions
p
erated protons at emission time of W ≈ 2.1×1050/n erg aboutthis.
p cm−3
woulddescribetheH.E.S.S.measurements,butisincompatible
withtheFermi-LATresults(seeFig.5blackdashedline).This
5. Conclusion
wasalreadypointedoutbyLemoine-Goumardetal.(2012).The
proton spectrum, which reproduces the γ-ray data, has a spec- Inthiswork,wepresentedthefirststrongevidencethatRCW86
tral index of Γp ≈ 1.7 and a cutoff energy of Ecut = 35 TeV. showsashell-typemorphologyinTeVγ−rays.TheTeVγ−ray
The index lies between the spectral index of the test particle spectrum favors an exponential cutoff power law with Γ =
approach and the spectral index with strong modified shocks 1.59±0.22 ±0.2 andacutoffenergyE =3.47±1.23 ±
stat sys cut stat
(Malkov1999;Berezhko&Ellison1999).ThebluelineinFig.5 2.2 TeVratherthanapurepowerlaw.
sys
presents this hadronic model. The total energy of all protons ThebroadbandSEDcanbewelldescribedbyasimplelep-
above 1 GeV is Wp ≈ 9×1049/ncm−3 erg, which means that a tonic one-zone model with a magnetic field strength of B ≈
fractionof0.1/ncm−3 ofthesupernova(TypeIa)energyhastobe 22µGandaΓ≈2.3power-lawelectronspectrumwithanexpo-
convertedintohigh-energyprotons.Theelectron-to-protonratio nential cutoff at E ≈ 19 TeV. The kinetic energy of all elec-
is Kep = 10−3. To model the radio and X-ray data, an electron trons above 1 GeVcutis about 0.1% of the kinetic energy of the
spectrumwithspectralindexofΓe ≈ 2.3andacutoffenergyof supernovaexplosionof1051erg.
E =9TeVisneededandamagneticfieldstrengthof100µG.
cut Modeling the broadband data using a hadronic one-zone
Thisiscomparabletotwodifferentestimates:onebyVo¨lketal. model requires a hard proton spectrum with index Γ ≈ 1.7.
(2005) of 99+−4266µG, which was deduced from the thickness of Thismodelisincompatiblewithaconventional∝ E−2pacceler-
the filaments in the SW region, and another one calculated by
ationspectrum,butliesbetweenthespectralindexvalueofthe
Arbutina et al. (2012) of a volume-averaged magnetic field of
test particle approach and the one with strong modified shocks
∼ 70 µG. The total kinetic energy of the protons above 1 GeV
(Malkov 1999; Berezhko & Ellison 1999). The total energy of
was found to be W ≈ 9×1049/n erg, which means that
p cm−3 all protons above 1 GeV is W ≈ 9 × 1049/n erg, when a
about 10%/n of the supernova (Type Ia) energy has to be p cm−3
cm−3 distanceof2.5kpcareassumed.Asaresultoflimitedstatistics
convertedintohigh-energyprotons.
neithertheleptonicnorthehadronicmodelcanberuledout.
Asalreadymentioned,thehadronicmodelcannotreproducethe
radio data with an electron spectrum that has the same spectral
Acknowledgements. The support of the Namibian authorities and of the
indexastheprotonspectrum(Γp =1.7).Thisisinconflictwith UniversityofNamibiainfacilitatingtheconstructionandoperationofH.E.S.S.
5
H.E.S.S.Collaborationetal.:DetailedanalysisoftheshelltypeSNRRCW86
eV] 102 MOST Leptonic Model
-1 s ATCA IC
-2 m ASCA IC CMB
c IC dust
x [ RXTE IC stars
u 10
Fl Fermi/LAT p0 decay
2 E H.E.S.S.
1
10-1
10-6 10-4 10-2 1 102 104 106 108 1010 1012 1014
E [eV]
Fig.4.SpectralenergydistributionofRCW86foraleptonicscenario.Theredsolidlinesdenotethetotalbroadbandemissionfrom
theone-zonemodelingdiscussedinSect.4ThedottedlinesshowtheIC-contributionsandthedashedlineisthatoftheπ0-decay
contribution.TheradiodatapointsarefromMolongloat408MHzandParkesat5GHz(Caswelletal.1975;Lemoine-Goumard
et al. 2012). X-ray data are from ASCA and RXTE from Lemoine-Goumard et al. (2012). The Fermi-LAT data points are taken
fromYuanetal.(2014)andtheH.E.S.S.dataarefromthisanalysis.
eV] 102 MOST Hadronic Model
-1 s ATCA IC
-2 m ASCA p0 decay
x [c RXTE G p = -2
u 10
Fl Fermi/LAT
2 E H.E.S.S.
1
10-1
10-6 10-4 10-2 1 102 104 106 108 1010 1012 1014
E [eV]
Fig.5. Spectral energy distribution of RCW 86 for a hadronic scenario. The solid blue lines denote the total broadband emission
fromtheone-zonemodelingdiscussedinSect.4ThedottedlinesshowtheIC-contributionsandthedashedlineisthatoftheπ0-
decay contribution. The black dashed dotted line shows the results for a proton spectrum with Γ = 2. The radio data points are
p
fromMolongloat408MHzandParkesat5GHz(Caswelletal.1975;Lemoine-Goumardetal.2012).X-raydataarefromASCA
and RXTEfrom Lemoine-Goumardet al.(2012). The Fermi-LATdata points aretaken fromYuan etal. (2014) andthe H.E.S.S.
dataarefromthisanalysis.
is gratefully acknowledged, as is the support by the German Ministry for fromtheAustralianResearchCouncilandtheScienceFoundationforPhysics
EducationandResearch(BMBF),theMaxPlanckSociety,theGermanResearch withinTheUniversityofSydney.
Foundation(DFG),theFrenchMinistryforResearch,theCNRS-IN2P3andthe
AstroparticleInterdisciplinaryProgrammeoftheCNRS,theU.K.Scienceand
TechnologyFacilitiesCouncil(STFC),theIPNPoftheCharlesUniversity,the
CzechScienceFoundation,thePolishMinistryofScienceandHigherEducation,
References
theSouthAfricanDepartmentofScienceandTechnologyandNationalResearch
Foundation, and by the University of Namibia. We appreciate the excellent
Acero, F., Aharonian, F., Akhperjanian, a. G., et al. 2010, Astronomy &
work of the technical support staff in Berlin, Durham, Hamburg, Heidelberg,
Astrophysics,516,7
Palaiseau,Paris,Saclay,andinNamibiaintheconstructionandoperationofthe
Aharonian, F., Akhperjanian, A. G., de Almeida, U. B., et al. 2009, The
equipment.TheMOSTisoperatedbyTheUniversityofSydneywithsupport
AstrophysicalJournal,692,1500
Aharonian,F.A.2013,AstroparticlePhysics,43,71
6
H.E.S.S.Collaborationetal.:DetailedanalysisoftheshelltypeSNRRCW86
Arbutina, B., Urosˇevic´, D., Andjelic´, M. M., Pavlovic´, M. Z., & Vukotic´, B.
2012,TheAstrophysicalJournal,746,79 1 Universita¨t Hamburg, Institut fu¨r Experimentalphysik, Luruper
Bamba,A.,Koyama,K.,&Tomida,H.2000,PublicationsoftheAstronomical
Chaussee149,D22761Hamburg,Germany
SocietyofJapan,52,1157
Berezhko,E.G.&Ellison,D.C.1999,TheAstrophysicalJournal,526,385 2 Max-Planck-Institut fu¨r Kernphysik, P.O. Box 103980, D 69029
Berge,D.,Funk,S.,&Hinton,J.2007,AstronomyandAstrophysics,466,1219 Heidelberg,Germany
Bernlo¨hr,K.,Carrol,O.,Cornils,R.,etal.2003,AstroparticlePhysics,20,111 3 DublinInstituteforAdvancedStudies,31FitzwilliamPlace,Dublin
Broersen,S.,Chiotellis,A.,Vink,J.,&Bamba,A.2014,MonthlyNoticesofthe 2,Ireland
RoyalAstronomicalSociety,441,3040 4 National Academy of Sciences of the Republic of Armenia,
Caswell,J.L.,Clark,D.H.,&Crawford,D.F.1975,Aust.J.Phys.,37,39 Marshall Baghramian Avenue, 24, 0019 Yerevan, Republic of
deLosReyesLopez,R.2011,inProceedingsofthe32ndInternationalCosmic
Armenia
RayConference(ICRC2011),216
5 Yerevan Physics Institute, 2 Alikhanian Brothers St., 375036
deNaurois,M.&Rolland,L.2009,AstroparticlePhysics,32,231
Yerevan,Armenia
Dickel,J.R.,Strom,R.G.,&Milne,D.K.2001,TheAstrophysicalJournal,
546,447 6 Institutfu¨rPhysik,Humboldt-Universita¨tzuBerlin,Newtonstr.15,
Ghavamian, P., Raymond, J., Smith, R. C., & Hartigan, P. 2001, The D12489Berlin,Germany
AstrophysicalJournal,547,995 7 UniversityofNamibia,DepartmentofPhysics,PrivateBag13301,
Helder,E.A.,Vink,J.,Bamba,A.,etal.2013,MonthlyNoticesoftheRoyal Windhoek,Namibia
AstronomicalSociety,435,910 8 GRAPPA, Anton Pannekoek Institute for Astronomy, University
Helder,E.A.,Vink,J.,Bassa,C.G.,etal.2009,Science,325,719 of Amsterdam, Science Park 904, 1098 XH Amsterdam, The
Hinton, J. A. & Hofmann, W. 2009, Annual Review of Astronomy &
Netherlands
Astrophysics,47,523
9 Department of Physics and Electrical Engineering, Linnaeus
Kelner,S.R.,Aharonian,F.A.,&Bugayov,V.V.2006,PhysicalReviewD,74,
University,35195Va¨xjo¨,Sweden
034018
Kesteven,M.J.&Caswell,J.L.1987,AstronomyandAstrophysics(ISSN0004- 10 Institut fu¨r Theoretische Physik, Lehrstuhl IV: Weltraum und
6361),183,118 Astrophysik, Ruhr-Universita¨t Bochum, D 44780 Bochum,
Lemoine-Goumard, M., Renaud, M., Vink, J., et al. 2012, Astronomy and Germany
Astrophysics,545,28 11 GRAPPA,AntonPannekoekInstituteforAstronomyandInstituteof
Long,K.S.&Blair,W.P.1990,AstrophysicalJournal,358,L13 High-EnergyPhysics,UniversityofAmsterdam,SciencePark904,
Malkov,M.A.1999,TheAstrophysicalJournal,511,L53 1098XHAmsterdam,TheNetherlands
Ohm,S.,vanEldik,C.,&Egberts,K.2009,AstroparticlePhysics,31,383 12 Institut fu¨r Astro- und Teilchenphysik, Leopold-Franzens-
Pisarski,R.L.,Helfand,D.J.,&Kahn,S.M.1984,AstrophysicalJournal,277,
Universita¨tInnsbruck,A-6020Innsbruck,Austria
710
13 School of Chemistry & Physics, University of Adelaide, Adelaide
Porter,T.A.&Strong,A.W.2005,Proc.29thICRC,4,77
Rho, J., Dyer, K. K., Borkowski, K. J., & Reynolds, S. P. 2002, The 5005,Australia
AstrophysicalJournal,581,1116 14 Centre for Space Research, North-West University, Potchefstroom
Smith,R.C.1997,AstronomicalJournalv.114,114,2664 2520,SouthAfrica
Stephenson,F.R.&Green,D.A.2002,Historicalsupernovaeandtheirrem- 15 LUTH, Observatoire de Paris, CNRS, Universite´ Paris Diderot, 5
nants,5 PlaceJulesJanssen,92190Meudon,France
Vink,J.,Bleeker,J.,vanderHeyden,K.,etal.2006,TheAstrophysicalJournal, 16 LPNHE,Universite´ PierreetMarieCurieParis6,Universite´ Denis
648,L33 Diderot Paris 7, CNRS/IN2P3, 4 Place Jussieu, F-75252, Paris
Vink,J.,Kaastra,J.S.,&Bleeker,J.A.M.1997,AstronomyandAstrophysics,
Cedex5,France
328,628
17 Laboratoire Univers et Particules de Montpellier, Universite´
Vo¨lk, H. J., Berezhko, E. G., & Ksenofontov, L. T. 2005, Astronomy and
Astrophysics,433,229 Montpellier 2, CNRS/IN2P3, CC 72, Place Euge`ne Bataillon, F-
Whiteoak, J. B. Z. & Green, A. J. 1996, Astronomy and Astrophysics 34095MontpellierCedex5,France
Supplement,118,329 18 DSM/Irfu,CEASaclay,F-91191Gif-Sur-YvetteCedex,France
Williams, B. J., Blair, W. P., Blondin, J. M., et al. 2011, The Astrophysical 19 Astronomical Observatory, The University of Warsaw, Al.
Journal,741,96 Ujazdowskie4,00-478Warsaw,Poland
Yamaguchi, H., Koyama, K., & Uchida, H. 2011, Publications of the 20 AixMarseilleUniversie´,CNRS/IN2P3,CPPMUMR7346,13288
AstronomicalSocietyofJapan,63,837
Marseille,France
Yuan, Q., Huang, X., Liu, S., & Zhang, B. 2014, The Astrophysical Journal
21 Instytut Fizyki Ja¸drowej PAN, ul. Radzikowskiego 152, 31-342
Letters,785,L22
Krako´w,Poland
22 Funded by EU FP7 Marie Curie, grant agreement No. PIEF-GA-
2012-332350,
23 School of Physics, University of the Witwatersrand, 1 Jan Smuts
Avenue,Braamfontein,Johannesburg,2050SouthAfrica
24 Laboratoire d’Annecy-le-Vieux de Physique des Particules,
Universite´ SavoieMont-Blanc,CNRS/IN2P3,F-74941Annecy-le-
Vieux,France
25 Landessternwarte, Universita¨t Heidelberg, Ko¨nigstuhl, D 69117
Heidelberg,Germany
26 Universite´Bordeaux,CNRS/IN2P3,Centred’E´tudesNucle´airesde
BordeauxGradignan,33175Gradignan,France
27 OskarKleinCentre,DepartmentofPhysics,StockholmUniversity,
AlbanovaUniversityCenter,SE-10691Stockholm,Sweden
28 WallenbergAcademyFellow,
29 Institut fu¨r Astronomie und Astrophysik, Universita¨t Tu¨bingen,
Sand1,D72076Tu¨bingen,Germany
30 LaboratoireLeprince-Ringuet,EcolePolytechnique,CNRS/IN2P3,
F-91128Palaiseau,France
31 APC, AstroParticule et Cosmologie, Universite´ Paris Diderot,
CNRS/IN2P3, CEA/Irfu, Observatoire de Paris, Sorbonne Paris
7
H.E.S.S.Collaborationetal.:DetailedanalysisoftheshelltypeSNRRCW86
Cite´, 10, rue Alice Domon et Le´onie Duquet, 75205 Paris Cedex
13,France
32 Univ.GrenobleAlpes,IPAG,F-38000Grenoble,France
CNRS,IPAG,F-38000Grenoble,France
33 DepartmentofPhysicsandAstronomy,TheUniversityofLeicester,
UniversityRoad,Leicester,LE17RH,UnitedKingdom
34 NicolausCopernicusAstronomicalCenter,ul.Bartycka18,00-716
Warsaw,Poland
35 Institut fu¨r Physik und Astronomie, Universita¨t Potsdam, Karl-
Liebknecht-Strasse24/25,D14476Potsdam,Germany
36 DESY,D-15738Zeuthen,Germany
37 Obserwatorium Astronomiczne, Uniwersytet Jagiellon´ski, ul. Orla
171,30-244Krako´w,Poland
38 Universita¨t Erlangen-Nu¨rnberg, Physikalisches Institut, Erwin-
Rommel-Str.1,D91058Erlangen,Germany
39 Centre for Astronomy, Faculty of Physics, Astronomy and
Informatics, Nicolaus Copernicus University, Grudziadzka 5, 87-
100Torun,Poland
40 DepartmentofPhysics,UniversityoftheFreeState,POBox339,
Bloemfontein9300,SouthAfrica
41 HeisenbergFellow(DFG),ITAUniversittHeidelberg,Germany,
42 GRAPPA, Institute of High-Energy Physics, University of
Amsterdam, Science Park 904, 1098 XH Amsterdam, The
Netherlands
e-mail:[email protected]
8