aa r X i v : . [ a s t r o - ph ] D ec Astronomy&Astrophysicsmanuscript no. casamag c (cid:13)
ESO 2018October 26, 2018
Observation of VHE γ -rays from Cassiopeia A with the MAGICtelescope J. Albert , E. Aliu , H. Anderhub , P. Antoranz , A. Armada , C. Baixeras , J. A. Barrio H. Bartko , D. Bastieri ,J. K. Becker , W. Bednarek , K. Berger , C. Bigongiari , A. Biland , R. K. Bock , , P. Bordas , V. Bosch-Ramon ,T. Bretz , I. Britvitch , M. Camara , E. Carmona , A. Chilingarian , J. A. Coarasa , S. Commichau , J. L. Contreras ,J. Cortina , M.T. Costado , V. Curtef , V. Danielyan , F. Dazzi , A. De Angelis , C. Delgado , R. de los Reyes ,B. De Lotto , E. Domingo-Santamar´ıa , D. Dorner , M. Doro , M. Errando , M. Fagiolini , D. Ferenc ,E. Fern´andez , R. Firpo , J. Flix , M. V. Fonseca , L. Font , M. Fuchs , N. Galante , R. Garc´ıa-L ´opez ,M. Garczarczyk , M. Gaug , M. Giller , F. Goebel , D. Hakobyan , M. Hayashida , T. Hengstebeck , A. Herrero ,D. H ¨ohne , J. Hose , C. C. Hsu , P. Jacon , T. Jogler , R. Kosyra , D. Kranich , R. Kritzer , A. Laille , E. Lindfors ,S. Lombardi , F. Longo , J. L ´opez , M. L ´opez , E. Lorenz , , P. Majumdar , G. Maneva , K. Mannheim ,O. Mansutti , M. Mariotti , M. Mart´ınez , D. Mazin , C. Merck , M. Meucci , M. Meyer , J. M. Miranda ,R. Mirzoyan , S. Mizobuchi , A. Moralejo , K. Nilsson , J. Ninkovic , E. O ˜na-Wilhelmi ⋆ , N. Otte , I. Oya ,D. Paneque , M. Panniello , R. Paoletti , J. M. Paredes , M. Pasanen , D. Pascoli , F. Pauss , R. Pegna ,M. Persic , , L. Peruzzo , A. Piccioli , M. Poller , N. Puchades , E. Prandini , A. Raymers , W. Rhode , M. Rib´o ,J. Rico , M. Rissi , A. Robert , S. R ¨ugamer , A. Saggion , A. S´anchez , P. Sartori , V. Scalzotto , V. Scapin ,R. Schmitt , T. Schweizer , M. Shayduk , , K. Shinozaki , S. N. Shore , N. Sidro , A. Sillanp¨a¨a , D. Sobczynska ,A. Stamerra , L. S. Stark , L. Takalo , P. Temnikov , D. Tescaro , M. Teshima , N. Tonello , D. F. Torres ,N. Turini , H. Vankov , V. Vitale , R. M. Wagner , T. Wibig , W. Wittek , F. Zandanel , R. Zanin , and J. Zapatero (A ffi liations can be found after the references) Received / Accepted
ABSTRACT
Aims.
We searched for very high energy (VHE) γ -ray emission from the supernova remnant Cassiopeia A Methods.
The shell-type supernova remnant Cassiopeia A was observed with the 17 meter MAGIC telescope between July 2006 and January 2007for a total time of 47 hours.
Results.
The source was detected above an energy of 250 GeV with a significance of 5.2 σ and a photon flux above 1 TeV of (7.3 ± stat ± sys ) × − cm − s − . The photon spectrum is compatible with a power law dN / dE ∝ E − Γ with a photon index Γ= ± stat ± sys . The source is point-likewithin the angular resolution of the telescope. Key words. acceleration of particles - ISM: cosmic rays - gamma rays: observations - ISM: supernova remnants - gamma rays: individual objects:Cassiopeia A
1. Introduction
Cassiopeia A (Cas A), with right ascension (RA) and declination(DEC) (23.385 h ,58.800 o ), is a prominent shell type supernovaremnant and a bright source of synchrotron radiation observed atradio frequencies, see Bell et al. (1975); Tu ff s et al. (1986), andin the X-ray band, see Allen et al. (1997), Favata et al. (1997).The remnant results from the youngest known Galactic super-nova, whose explosion took place around 1680. Its distance wasestimated at 3.4 kpc by Reed et al. (1995). High resolution X-ray images from the Chandra satellite, see Hughes et al. (2000),reveal a shell-type nature of the remnant and the existence of acentral object. The progenitor of Cas A was probably a Wolf-Rayet star, as discussed in Fesen et al. (1991) and Iyudin et al. Send o ff print requests to :E. O˜na-Wilhelmi, e-mail: [email protected]. Vitale e-mail: [email protected] ⋆ Present address:
APC, Paris, France (1997). The progenitor’s initial mass was large, estimated to bebetween 15 and 25 M ⊙ , see Young et al. (2006). The morphol-ogy of the remnant as seen in optical, X-ray and IR wavelengthconsists on a patchy and irregular shell with a diameter of 4’(4 pc at 3.4 kpc). The supernova blast wave is expanding into awind bubble formed from the previous wind phases of the pro-genitor star; this plays an important role in shock acceleration ofCR, see Berezhko et al. (2003).At TeV energies, Cas A was detected by the HEGRAStereoscopic Cherenkov Telescope System, which accumu-lated 232 hours of data from 1997 to 1999. TeV γ -ray emission was detected at 5 σ level and a flux of(5.8 ± stat ± sys ) 10 − ph cm − s − above 1 TeV was derived,as discussed in Aharonian et al. (2001). The spectral distributionbetween 1 and 10 TeV was found to be consistent with a powerlaw with a di ff erential spectral index of -2.5 ± stat ± sys .Upper limits at TeV energies have been set also by Whipple, seeLessard et al. (1999) and CAT, see Goret et al. (1999). These up- Albert et al: Observation of VHE γ -rays from Cassiopeia A with the MAGIC telescope per limits were consistent with the HEGRA detected flux level.At lower energy, EGRET set an upper limit for a flux below12 . × − cm − s − , see Esposito et al. (1996).The HEGRA detection makes Cas A a good scenario to testthe supernova remnant emission at lower energies, in particularfor trying to further distinguish between leptonic and hadronicmodels for the origin of the γ -ray emission. A summary of theobservations and analysis results is given in Section 2, the resultsare reported in Section 3 and finally a comparison of MAGICdetection with the existing model predictions for the TeV γ -rayemission on Cas A is discussed in Section 4.
2. Observations and Data Analysis
The MAGIC (Major Atmospheric Gamma Imaging Cherenkov)Telescope is located on the Canary Island La Palma (2200 masl, 28 ◦ ′ N , 17 ◦ ′ W ) and has a 17 m-diameter tessellated re-flector dish, see Cortina et al. (2005). The total field of view is3.5 ◦ . The accessible energy range spans from 50-60 GeV (triggerthreshold at low zenith angle) up to tens of TeV. The telescopeangular resolution (sigma of the Gaussian fit to the point spreadfunction, PSF, σ ps f ) is about 0.09 ◦ .Cas A observations were performed between June 2006 andJanuary 2007 for a total observation time of 47 hours after qual-ity cuts, namely, after rejecting runs with detector problems oradverse atmospheric conditions. The zenith angle ranged from29 ◦ to 45 ◦ and averaged 35 ◦ . The observation technique ap-plied was the so-called wobble mode, see Daum et al. (1997)in which the telescope pointed alternatively for 20 minutes totwo opposite sky positions at 0.4 ◦ o ff the source. Most of thedata were taken under moderate moonlight illumination (86%of the scheduled observation time). Depending on the di ff erentmoonlight levels, the resulting PMT anode currents ranged be-tween 1 µ A and 6 µ A, as compared to a typical anode currentof 1 µ A for dark night observations. Correspondingly, the trig-ger discriminator threshold (DT) was varied between 15 and 30arbitrary units (a.u.) to keep a low rate of accidental events.The mean trigger discriminator threshold during the observa-tions was 19 a.u, which corresponds to 13.3 photoelectrons (PE).Briefly, the impact of the rise of DT can be summarized bya decrease on the relative γ -ray e ffi ciency from 1 (dark obser-vations) to 0.84 while the relative sensitivity worsens from2.5 to 2.7% with respect to the Crab flux. Although this ef-fect is important for images containing a low number of PE’s(low size), the energy threshold rise ( ∼ γ / hadronshower separation, the shower images were parameterized usingthe Hillas parameters, see Hillas et al. (1985). These variableswere combined for γ / hadron separation by means of a RandomForest classification algorithm, see Bock et al. (2005), trainedwith MC simulated γ -ray events and data from galactic areasnear the source under study but containing no γ -ray sources. TheRandom Forest method calculates for every event a parameterdubbed HADRONNESS (H), which parameterizes the purity of Minimal flux detectable with σ significance in 50 hours of obser-vations. E ve n t s ] [deg θ -50050100150 Convolved radial Gaussian FitMAGIC PSF
Fig. 1: The upper panel shows the distributions of θ (measured indegree ) in the direction of the source (black dots) and anti-sources(blue shaded histogram). A lower SIZE cut of 400 PE was applied. Thelower panel shows the θ distribution after background subtraction. Thevertical line shows the optimum angular cut. The red function corre-sponds to the telescope PSF and is derived from Crab nebula data takenin the same observational conditions as those for Cas A. The black dis-tribution is the result of a Gaussian fit to the excess distribution. hadron-initiated images in the multi-dimensional space definedby the Hillas variables.The θ distribution is computed for the source position,where θ is the angular distance between the source position inthe sky and the reconstructed origin position of the shower. Thereconstruction of individual γ -ray arrival directions makes use ofthe so-called DISP method (Domingo-Santamaria et al. (2005)).The expected number of background events are calculated usingfive regions symmetrically distributed for each wobble positionwith respect to the center of the camera and refered to as anti-sources.The optimum H and the angular cuts were derived using darknight Crab data of the same epoch and in the same observationconditions (zenith range, astronomical nights). The use of a darknight data sample in optimizing the telescope sensitivity is justi-fied by the results in Albert et al. (2007a). For the spectral anal-ysis, the energy of each individual γ -ray candidate was also es-timated using the Random Forest technique. The average energyresolution for the analyzed energy range was 20%.
3. Source detection, extension and energyspectrum
The so-called θ distributions for the source and anti-source po-sitions are shown in Figure 1 for a lower SIZE cut of 400 PE,which optimizes the MAGIC signal to noise ratio. The blackpoints correspond to the source position whereas the blue-shadedhistogram corresponds to the anti-sources. The subtraction of thetwo histograms shows the excess in the direction of Cas A. Anexcess of N excess =
157 with a significance of 5.2 σ (using thelikelihood method of Li & Ma (1983)) is detected within the re-gion 0.13 ◦ centered at the HEGRA position.Figure 2 shows the excess map of γ -ray candidates withimages larger than 400 PE. The map has been smeared with aGaussian of σ = o . The source position has been determinedby ways of a fit of the non-smeared sky map to a bidimensional lbert et al: Observation of VHE γ -rays from Cassiopeia A with the MAGIC telescope 3 -200204060 RA (J2000) [hrs]23.323.3523.423.45 D ec ( J2000 ) [ d e g ] PSF s m oo t h e d excesses Fig. 2: Sky map around the position of Cas A. A lower cut in SIZE of400 PE was applied. The green crosses mark the 2 wobble positions.The red cross indicates the MAGIC best fit position. The black crossmarks the HEGRA source position, which is within 1 standard deviationfrom the MAGIC one. The bars of the crosses for both the MAGIC andHEGRA marks correspond to 1 sigma statistical errors.
Gaussian function. The best fit position coordinates are RA = ± stat ± sys h and DEC = ± stat ± ◦ sys (for more details on the systematic uncertainties in the sourceposition determination, see Bretz et al. (2005)).In X-rays and radio-frequencies Cas A has an angular diame-ter of 0.08 ◦ , which is just on the limit of the MAGIC angular res-olution. The MAGIC system PSF is derived from MC simulationfor a point source, and is found to be σ ps f = ± ◦ (shownin Figure 2). This value was validated with Mkn 421 and CrabNebula data (see Albert et al. (2007b)). To further constrain theextension of the source we fit the excess with a Gaussian func-tion convolved with the PSF (F = A · exp ( − . θ / ( σ src + σ ps f ))).We obtain a value of σ src which is compatible with zero withinthe fit error. Figure 1 shows the telescope PSF and the result ofthe Gaussian fit (dotted blue curve).Figure 3 shows the reconstructed spectrum above 250 GeV.The spectrum is consistent with a power law (dN / dE ∝ E − Γ ).The di ff erential flux at 1 TeV is (1.0 ± stat ± sys ) × − TeV − cm − s − with a photon index of Γ= ± stat ± sys .Thesystematic error is estimated to be 35% in the flux level deter-mination and 0.2 in the spectral index (see Albert et al. (2006)).The measured spectrum was unfolded using the Gauss-Newtonmethod, see Schmelling (1994). The χ / d.o.f of the fit is 2.83 / σ error limit on the flux fitted is also added as a grey band.The Cas A flux corresponds to an integral flux above 1 TeV of3% of the Crab nebula flux above the same threshold (in reddashed line in figure 3, see Albert et al. (2007b)). The Cas Aspectrum measured by HEGRA is also shown as a blue solidline. The spectrum measured about 8 years later by MAGIC isconsistent with that measured by HEGRA for the energies above1 TeV, i.e, where they overlap.
4. Discussion
The VHE MAGIC 47-hour observation of Cas A confirms thesource detection by HEGRA after a multi-year integration of232 hours and at the same time significantly extends the en-ergy spectrum down to about 250 GeV. Cas A is detected withmore than 5 σ at a flux level compatible with the HEGRA mea-surement for those energies explored in common. The difer-ential flux at 1 TeV measured by MAGIC is 1.0 ± stat × − T eV − cm − s − to be compared with the one measured byHEGRA, 0.9 ± stat × − T eV − cm − s − . The agreement be- Energy [GeV] ] - T e V - s - D i ff e r e n t i a l f l u x [ c m -15 -14 -13 -12 -11 -10 Cas A MAGICCrab NebulaCas A HEGRA
Fig. 3: Cas A spectrum above 250 GeV. The blue line represents the ear-lier measurement by HEGRA. The red line represents the Crab nebulaspectrum. The shaded area is the 1 σ statistical error of the fit. tween the two measures is excellent not only in the determina-tion of the flux level but also in the spectral index measured.Although the errors in the spectral index are large, there is no ev-idence for a high energy cuto ff , nor for a deviation from a powerlaw at lower energies. The detection of very high energy γ -rays from Cas A provides evidence of the acceleration of multi-TeV particles in SNR shocks and their visibility in gamma-raysDrury et al. (1994).Significant e ff orts have been made for the theoretical mod-eling of Cas A’s multi-frequency emission, including that at thehighest energies. The e ff ect of an energy-dependent propagationof relativistic electrons in a spatially inhomogeneous mediumhas been used in order to interpret the radio emission from theregion and define its electron content (Atoyan et al. (2000a)).The variations in brightness in the radio band is so complex thata multi-zone model was used: distinguishing between compact,bright spectrum radio knots and the bright fragmented radio ringon one hand, and a di ff use plateau on the other. A three-zonemodel with a magnetic field decreasing from its highest valuein the compact zones putatively related with acceleration sites,to a lower value in regions surrounding the shell, to yet a lowervalue in the neighborhood has been found to reproduce the radiodata, with a magnetic field around and below 1 mG. The fluxesat TeV energies, due to Bremsstrahlung and inverse Comptonradiation of the same relativistic electrons have also been com-puted (Atoyan et al. (2000b)) and, albeit the parameters allow alarge range of possible fluxes, the overall shape of the spectrumremains similar, showing a steep cuto ff for multi-TeV energies(see, e.g., Figure 7 of Atoyan et al. (2000b)). This cuto ff is notseen in HEGRA and / or MAGIC data, disfavoring a leptonic ori-gin of the radiation. Vink and Laming (2003) also studied multi-zone models for Cas A, assuming no di ff erence between zonesother than in their magnetic field. They found that an IC originof high energy fluxes would be possible but only for low val-ues (within the range allowed to be consistent with radio and X-ray observations, see e.g., Vink and Laming 2003, Hwang et al.2004) of the magnetic field and high, far-infrared photon density.The generally high values of the magnetic field necessary to ex-plain the multi-frequency observations makes it likely that TeVemission from Cas A is then dominated by pion decay (Atoyanet al. 2000b, Vink and Laming 2003).Berezhko et al. (2003) applied a non-linear kinetic model ofcosmic-ray acceleration to describe Cas A, ignoring the role ofany small scale inhomogeneities for the production of the veryhigh energy particles and considering the whole SNR blast wave Albert et al: Observation of VHE γ -rays from Cassiopeia A with the MAGIC telescope log(Energy) [eV] ] - s - E n e r g y * F l u x [ T e V c m -15 -14 -13 -12 -11 o π NBIC
Crab NebulaHEGRA IACT SystemMAGIC 2006EGRETWhippleCAT
Fig. 4: Spectra of Cas A as measured by MAGIC. The shaded areaaround the 0.65 TeV detection shows the 1 σ statistical error range un-der the assumption of a E − α power law spectrum. The upper limits givenby Whipple, EGRET and CAT are also indicated, as well as the HEGRAdetection. The MAGIC and HEGRA spectra are shown in the contextof the model by Berezhko et al. (2003). Both hadronic ( π o with andwithout an energy cuto ff ) and leptonic (NB and IC) γ -ray emission areshown. The normalization of the pion decay spectrum can be taken as afree parameter. as the main relativistic particle generator. Figure 4 representsthe expected integral γ -ray flux components from non-thermalBremstrahlung NB, IC scattering on the background radiationfield (cosmic microwave + optical / infrared), and hadronic colli-sions of CR protons with gas nuclei, respectively, for this model.The pion-decay γ -ray flux presented in Figure 4 –with and with-out an exponential cuto ff at 4 TeV– was calculated with a renor-malization factor of 1 / ffi ciently injects and accelerates cosmicrays). This emphasizes that the normalization of nucleonic pre-dictions of γ -rays is to be considered a free parameter, withincertain reasonable boundaries. The predicted slope for the dom-inating nucleonic-produced γ -rays (that dominates, even whenall possible uncertainties leading to an increase of the leptonicemission are included) is hard in the range of interest, as shownin Figure 4, perhaps too hard already to provide a good fit tothe new MAGIC data at low energies. Higher and lower energymeasurements, and a better signal to noise ratio for the spectrumdetermination of such a weak source, are still needed for a defi-nite answer. Acknowledgements.
We would like to thank the IAC for the excellent work-ing conditions at the Observatorio del Roque de los Muchachos in La Palma.The support of the German BMBF and MPG, the Italian INFN and the SpanishCICYT is gratefully acknowledged. This work was also supported by ETHResearch Grant TH 34 /
04 3 and the Polish MNiI Grant 1P03D01028.
References
Aharonian, F. et al. 2001, A&A 112, 307.Albert, J. et al., (MAGIC Collaboration), “Very high energy gamma-ray obser-vations during moonlight and twilight with the MAGIC telescope”, submittedto ApJ, astro-ph / / ff s, R.J., & V¨ o lk, H.J., 2000a, A&A 354, 915.Atoyan, A.M., Tu ff s, R.J., & V¨ o lk, H.J., 2000b, A&A 355, 211.Bell, A.R., Gull, S.F. & Kenderline, S., 1975, Nature 257, 463.Berezhko, E.G, P¨ u hlhofer & V¨ o lk, H.J. 2003, A&A 400, 971. Bretz, T. et al., (MAGIC Collaboration) 2003, Proc. of 28th ICRC (Tsukuba),2947.Bock, R. K. et al. 2005, Nucl. Instrum. Meth. A516, 188.Cortina, J. et al., (MAGIC Collaboration), AIP Conf. Proc. 2005 745, 730.Daum, A. et al., 1997, Astrop. Physics 8, 1.Domingo-Santamaria, E. et al., 2005, Proc. of 29th ICRC (Pune) 5, 363.Drury L.O’C. et al., 1994 , A&A 287, 959.Esposito, J.A. et al., 1996, ApJ, 461, 820.Favata, F., Vink, J., Fiume, D.D., et al., 1997, A&A 324, L49.Fesen, R.A. & Becker, R.H., 1991, ApJ, 371, 621.Gaug, M. et al., 2005, Proc. 29th ICRC (Pune) 5, 375.Goret, P., Goui ff es, C., Nuss, E. & Ellison, D.C., 1999, in Proc. of the 26th ICRC,ed. D. Kieda, M. Salamon & B. Dingus (Salt Lake City), vol. 3, 496.Hillas, A.M., 1985, Proc. of 19th ICRC, La Joya 3, 445.Hughes, J.P., Rakowski, C.E., Burrows, D.N. & Slane, P.O., 2000, ApJ 528,L109.Hwang, U. et al., 2004, ApJ Letters 615, 117.Iyudin, A.F., Diehl, R., Lichti, G.G., et al. 1997, in 2nd INTEGRAL WorkshopThe Transparent Universe, St. Malo, France (ESA Publications Division,ESTEC, Noordwijk, The Netherlands), 37.Lessard, R.W., Bond, I.H., Boyle, P.J., et al. 1999, in Proc. of the 26th ICRC, ed.D. Kieda, M. Salamon & B. Dingus (Salt Lake City), vol. 3, 488.Li, T. & Ma, Y., 1983, ApJ 272, 317.Reed, J.E., Hester, J.J, Fabian, A.C. & Winkler, P.F., 1995, ApJ 440, 706.Schmelling, M., 1994, NIM A, 340, 400.Tu ff s, R.J., 1986, MNRAS 219, 13.Vink, J. & Laming, J. M. 2003, ApJ 584, 758.V¨olk, H. J., 1997, in Towards a Major Atmospheric Cherenkov DetectorV, 87,Space Research Unit, O. C. de Jagger Editor.Young, P.A et al., 2006, ApJ 640, 891 Universit¨at W¨urzburg, D-97074 W¨urzburg, Germany Institut de F´ısica d’Altes Energies, Edifici Cn., E-08193 Bellaterra(Barcelona), Spain ETH Zurich, CH-8093 Switzerland Universidad Complutense, E-28040 Madrid, Spain Universitat Aut`onoma de Barcelona, E-08193 Bellaterra, Spain Max-Planck-Institut f¨ur Physik, D-80805 M¨unchen, Germany Universit`a di Padova and INFN, I-35131 Padova, Italy Universit¨at Dortmund, D-44227 Dortmund, Germany University of Ł´od´z, PL-90236 Lodz, Poland Universitat de Barcelona, E-08028 Barcelona, Spain Yerevan Physics Institute, AM-375036 Yerevan, Armenia Tuorla Observatory, Turku University, FI-21500 Piikki¨o, Finland Instituto de Astrofisica de Canarias, E-38200, La Laguna,Tenerife, Spain Universit`a di Udine, and INFN Trieste, I-33100 Udine, Italy Universit`a di Siena, and INFN Pisa, I-53100 Siena, Italy University of California, Davis, CA-95616-8677, USA Humboldt-Universit¨at zu Berlin, D-12489 Berlin, Germany Institute for Nuclear Research and Nuclear Energy, BG-1784Sofia, Bulgaria INAF / Osservatorio Astronomico and INFN Trieste, I-34131Trieste, Italy Universit`a di Pisa, and INFN Pisa, I-56126 Pisa, Italy21