aa r X i v : . [ a s t r o - ph ] S e p Chin. J. Astron. Astrophys. Vol. 0, No. 0, (200x) 000–000( ) Chinese Journal ofAstronomy andAstrophysics
Observations of VHE γ -Ray Sources with the MAGICTelescope H. Bartko ⋆ for the MAGIC Collaboration Max-Planck-Institute for Physics, Munich, Germany
Received 2007 month day; accepted 2007 month day
Abstract
The MAGIC telescope with its 17m diameter mirror is today the largestoperating single-dish Imaging Air Cherenkov Telescope (IACT). It is located onthe Canary Island La Palma, at an altitude of 2200m above sea level, as part of theRoque de los Muchachos European Northern Observatory. The MAGIC telescopedetects celestial very high energy γ -radiation in the energy band between about50 GeV and 10 TeV. Since Autumn of 2004 MAGIC has been taking data routinely,observing various objects like supernova remnants (SNRs), γ -ray binaries, Pulsars,Active Galactic Nuclei (AGN) and Gamma-ray Bursts (GRB). We briefly describethe observational strategy, the procedure implemented for the data analysis, anddiscuss the results for individual sources. An outlook to the construction of thesecond MAGIC telescope is given. Key words:
TeV γ -ray astrophysics – super nova remnants, pulsars, binarysystems – AGN, blazars – EBL – GRBs – dark matter One of the most important ’messengers’ of many high energy phenomena in our universe are γ -rays. The detection of very high energy (VHE, E γ >
100 GeV) cosmic γ -radiation by ground-based Cherenkov telescopes has opened a new window to the Universe, called VHE γ -ray as-tronomy. It is a rapidly expanding field with a wealth of new results, particularly during thelast two years, due to the high sensitivity of a new generation of instruments. The major scien-tific objective of γ -ray astronomy is the understanding of the production, acceleration, trans-port and reaction mechanisms of VHE particles in astronomical objects. This is tightly linkedto the search for sources of the cosmic rays. The MAGIC (Major Atmospheric γ -ray ImagingCherenkov) telescope is one of the new generation of Imaging Air Cherenkov Telescopes (IACT)for VHE γ -ray astronomy.The physics program of the MAGIC telescope includes topics, both of fundamental physicsand astrophysics. This article is structured as follows: in section 2 the MAGIC telescope ispresented and the data analysis is explained. The main part of this paper reviews the resultsof observations with the MAGIC telescope: in section 3 the results of Galactic sources and in ⋆ E-mail: [email protected]
H. Bartko for the MAGIC collaboration section 4 the results of extragalactic sources are described. Sections 5 and 6 deal with the searchfor γ -ray emission from Gamma Ray Bursts (GRBs) and dark matter particle annihilation.Finally, section 7 contains the conclusions and an outlook to the second MAGIC telescope. MAGIC (Baixeras et al., 2004; Cortina et al., 2005) is currently the largest single-dish ImagingAir Cherenkov Telescope (IACT) in operation. Located on the Canary Island La Palma (28 . ◦ N,17 . ◦ W, 2200 m a.s.l.), it has a 17-m diameter tessellated parabolic mirror, supported by a light-weight carbon fiber frame. It is equipped with a high-quantum-efficiency 576-pixel 3 . ◦ field-of-view photomultiplier tube (PMT) camera. The analog signals are transported via optical fibersto the trigger electronics and were read-out by a 300 MSampels/s FADC system till March 2007and by a new 2 GSamples/s FADC system (Bartko et al., 2005a) thereafter.The MAGIC telescope can operate under moderate moonlight or twilight conditions(Albert et al., 2007j). For these conditions, no change in the high voltage settings is neces-sary as the camera PMTs were especially designed to avoid high currents.The data analysis is generally carried out using the standard MAGIC analysis and recon-struction software (Bretz & Wagner, 2003), the first step of which involves the FADC signalreconstruction and the calibration of the raw data (Gaug et al., 2005; Albert et al., 2006i). Aftercalibration, image-cleaning tail cuts are applied (see e.g. Fegan, 1997).The camera images areparameterized by image parameters (Hillas, 1985). The Random Forest method (see Bock et al.(2004); Breiman (2001) for a detailed description) was applied for the γ /hadron separation (fora review see e.g. Fegan, 1997) and the energy estimation.For each event, the arrival direction of the primary γ -ray candidate in sky coordinatesis estimated using the DISP-method resulting in VHE γ -ray sky maps (Fomin et al., 1994;Lessard et al., 2001; Domingo-Santamaria et al., 2005). The angular resolution of this procedureis ∼ . ◦ , while the source localization in the sky is provided with a systematic error of 1 ′ (Albert et al., 2007k).The differential VHE γ -ray spectrum (dN γ / (dE γ dAdt) vs. true E γ ) is corrected (unfolded)for the instrumental energy resolution (Albert et al., 2007o).All fits to the spectral points takeinto account the correlations between the spectral points that are introduced by the unfoldingprocedure. The systematic error in the flux level determination depends on the slope of the γ -ray spectrum. It is typically estimated to be 35% and the systematic error in the the spectralindex is 0.2 (Albert et al., 2006b, 2007k). The observations with the MAGIC telescope included the following types of objects: VHE γ -ray sources coincident with supernova remnants (section 3.1), the Galactic Center (section 3.2), γ -ray binaries (section 3.3), pulsars and pulsar wind nebulae (section 3.4). γ -ray sources coincident with Supernova remnants Shocks produced by supernova explosions are assumed to be the source of the galactic com-ponent of the cosmic ray flux (Baade & Zwicky, 1934). In inelastic collisions of high energycosmic rays with ambient matter γ -rays and neutrinos are produced. These neutral particlesgive direct information about their source, as their trajectories are not affected by the Galacticand extra Galactic magnetic fields in contrast to the charged cosmic rays. However, not all VHE γ -rays from galactic sources are due to the interactions of cosmic rays with ambient matter.There are also other mechanisms for the production of VHE γ -rays like the inverse Compton bservations of VHE γ -Ray Sources with the MAGIC Telescope 3 up-scattering of ambient low energy photons by VHE electrons. For each individual source ofVHE γ -rays, the physical processes of particle acceleration and γ -ray emission in this sourcehave to be determined. A powerful tool is the modeling of the multiwavelength emission of thesource taking into account the ambient gas density as traced by CO observations (Torres et al.,2003).Within its program of observation of galactic sources, MAGIC has taken data on a numberof supernova remnants, resulting in the discovery of VHE γ -ray emission from a source in theSNR IC443, MAGIC J0616+225 (Albert et al., 2007e). Moreover, two recently discovered VHE γ -ray sources, which are spatially coincident with SNRs, HESS J1813-178 and HESS J1834-087 (Aharonian et al., 2006a) have been observed with the MAGIC telescope (Albert et al.,2006a,g). Recently, also VHE γ -rays have been observed from the SNR Cas A (Albert et al.,2007l).
IC443 is a well-studied shell-type SNR near the Galactic Plane with a diameter of 45’ ata distance of about 1.5 kpc. It is a prominent source and it has been studied from radio wavesto γ -rays of energies around 1 GeV. Gaisser et al. (1998) as well as other groups extrapolatedthe energy spectrum of 3EG J0617+2238 into the VHE γ -ray range and predicted readilyobservable fluxes. Nevertheless, previous generation IACTs have only reported upper limits tothe VHE γ -ray emission (Khelifi, 2003; Holder, 2005). The observation of IC 443 using theMAGIC Telescope has led to the discovery of a new source of VHE γ -rays, MAGIC J0616+225.The flux level of MAGIC J0616+225 is lower and the energy spectrum (fitted with a power lawof slope Γ = − . ± .
3) is softer than the predictions (Gaisser et al., 1998). The coincidenceof the VHE γ -ray source with SNR IC 443 suggests this SNR as a natural counterpart. Amassive molecular cloud and OH maser emissions are located at the same sky position as thatof MAGIC J0616+225, see figure 2. This suggests that a hadronic origin of the VHE γ -rays ispossible. However, other mechanisms for the VHE γ -ray emission cannot be excluded yet. HESS J1834-087 is spatially coincident with the SNR G23.3-0.3 (W41). W41 is an asym-metric shell-type SNR, with a diameter of 27’ at a distance of ∼ γ -ray emission. As in the case of IC 443, the VHE γ -radiationof W41 is associated with a large molecular complex called ”[23,78]” (Dame et al., 1986), seefigure 1. Although the mechanism responsible for the VHE γ -radiation has not yet been clearlyidentified, it could be produced by high energy hadrons interacting with the molecular cloud. HESS J1813-178 is spatially coincident with SNR G12.8-0.0 with a diameter of 2’ at adistance of ∼ γ -ray sources are usually spatially correlatedwith SNRs. Nevertheless, the exact nature of the parent particles of the VHE γ -rays, theiracceleration (in SNR shocks or PWN), and the processes of γ -ray emission need (or still require)further study. The Galactic Center region contains many remarkable objects which may be responsible forhigh-energy processes generating γ -rays: A super-massive black hole, supernova remnants, can-didate pulsar wind nebulae, a high density of cosmic rays, hot gas and large magnetic fields.Moreover, the Galactic Center may appear as the brightest VHE γ -ray source from the an- H. Bartko for the MAGIC collaboration l [deg] b [ d e g ] -1.5-1-0.500.51 s r - excess eve n t s / -200204060802222.52323.52424.5 PSF
Fig. 1
Sky map of γ -ray candidate events(background subtracted) in the direction ofHESS J1834-087 for an energy threshold ofabout 250 GeV. The source is clearly ex-tended with respect to the MAGIC PSF (smallwhite circle). The two white stars denotethe tracking positions of the MAGIC tele-scope. Overlayed are CO emission contours(black) from Dame et al. (2001) and contoursof 90 cm VLA radio data from White et al.(2005) (green). The CO contours are at25/50/75 K km/s, integrated from 70 to 85km/s in velocity, the range that best de-fines the molecular cloud associated with W41.The contours of the radio emission are at0.04/0.19/0.34/0.49/0.64/0.79 Jy/beam, cho-sen for best showing both SNRs G22.7-0.2 andG23.3-0.3 at the same time. Clearly, there is nosuperposition with SNR G22.7-0.2. The cen-tral white circle denotes the source region inte-grated for the spectral analysis. (Albert et al.,2006g). l [deg] b [ d e g ] s r - excess eve n t s / -20020406080100120140160188.5189189.5 PSF
Fig. 2
Sky map of γ -ray candidate events(background subtracted) in the direction ofMAGIC J0616+225 for an energy thresholdof about 150 GeV. The cyan CO contours(Dame et al., 2001) are at 7 and 14 K km/s,integrated from -20 to 20 km/s in velocity, therange that best defines the molecular cloud as-sociated with IC 443. The green contours of20 cm VLA radio data (Condon et al., 1998)are at 5 mJy/beam, chosen for best showingboth the SNR IC 443. The purple Rosat X-raycontours (Asaoka & Aschenbach, 1994) are at700 and 1200 counts / 6 · − sr. The blackEGRET contours (Hartman, 1999) represent a68% and 95% statistical probability that a sin-gle source lies within the given contour. Thewhite star denotes the position of the pul-sar CXOU J061705.3+222127 (Olbert et al.,2001). The black dot shows the position of the1720 MHz OH maser (Claussen et al., 1997).The white circle shows the MAGIC PSF of σ = 0 . ◦ . (Albert et al., 2007e). nihilation of possible dark matter particles (Bartko et al., 2005b) of all proposed dark matterparticle annihilation sources.The Galactic Center was observed with the MAGIC telescope (Albert et al., 2006b) underlarge zenith angles, resulting in the measurement of a differential γ -ray flux, consistent witha steady, hard-slope power law between 500 GeV and about 20 TeV, with a spectral index ofΓ = − . ± .
2. This result confirms the previous measurements by the HESS collaboration. TheVHE γ -ray emission does not show any significant time variability; the MAGIC measurementsrather affirm a steady emission of γ -rays from the GC region on time scales of up to one year.The VHE γ -ray source is centered at (RA, Dec)=(17 h m s , -29 ◦ ′ ). The excess is point-like, its location is consistent with SgrA ∗ , the candidate PWN G359.95-0.04 as well as SgrAEast. The nature of the source of the VHE γ -rays has not yet been (or yet to be) identified. Thepower law spectrum up to about 20 TeV disfavours dark matter annihilation as the main origin bservations of VHE γ -Ray Sources with the MAGIC Telescope 5 of the detected flux, see also Aharonian et al. (2006d). The absence of flux variation indicatesthat the VHE γ -rays are rather produced in a steady object such as a SNR or a PWN, and notin the central black hole. γ -ray binaries The γ -ray binary system LS I +61 303 is composed of a B0 main sequence star with acircumstellar disc, i.e. a Be star, located at a distance of ∼ e = 0 . ± . γ -ray sky map is shown in figure 3. The datawere first divided into two different samples, around periastron passage (0.2-0.3) and at higher(0.4-0.7) orbital phases. No significant excess in the number of γ -ray events is detected aroundperiastron passage, whereas there is a clear detection (9.4 σ statistical significance) at laterorbital phases. Two different scenarios were discussed to explain this high energy emissions:the microquasar scenario where the γ -rays are produced in a radio-emitting jet; or the pulsarbinary scenario, where they are produced in the shock which is generated by the interaction ofa pulsar wind and the wind of the massive companion.Recently, an excess 4 . σ significance (after trial correction) of γ -ray candidate events overthe expected background was observed during 79 minutes of one night for the black holeX-ray binary (BHXB) Cygnus X-1 (Albert et al., 2007f). Moreover, VHE γ -ray emissionhas been observed from the high mass X-ray Binary LS 5039 by the H.E.S.S. collaboration(Aharonian et al., 2005). The Crab Nebula is a bright and steady emitter of GeV and TeV energies, and is therefore anexcellent calibration candle. This object has been observed intensively in the past, over a widerange of wavelengths.The energy domain between 10 and 100 GeV is of particular interest, as both the InverseCompton peak of the spectral energy distribution and the cut-off of the pulsed emission isexpected in this energy range.A significant amount of MAGIC’s observation time has been devoted to observing the CrabNebula, both for technical (because it is a strong and steady emitter) and astrophysical studies.A sample of 16 hours of selected data has been used to measure the energy spectrum between60 GeV and 9 TeV, and the result is shown in figure 4 (Albert et al., 2007k). Also, a search forpulsed γ -ray emission from the Crab Pulsar has been carried out. Figure 5 shows the derived(95% CL.) upper limits.Pulsed γ -ray emission was also searched for from the pulsar PSR B1951+32. A 95% CL. of4 . · − cm − − was obtained for the flux of pulsed γ -ray emission for E γ >
75 GeV andof 1 . · − cm − − for the steady emission for E γ >
140 GeV (Albert et al., 2007i).
The detection and characterization of VHE γ -ray emitting Active Galactic Nuclei (AGN) isone of the main goals of ground–based γ -ray astronomy. The observational results can be usedto explore the physics of the relativistic jets in AGNs, to understand the origin of the VHE γ -rays as well as to correlate with each other the fluxes of photons in different energy bands(optic, X-rays and γ -rays). Moreover, perform population studies of AGNs can be performed,information about the extragalactic background light (EBL) density can be extracted and even H. Bartko for the MAGIC collaboration
Fig. 3
Smoothed maps of γ -ray excess events above 400 GeV around LS I +61 303.(A) 15.5 hours corresponding to data around periastron, i.e. between orbital phases0.2 and 0.3. (B) 10.7 hours at orbital phase between 0.4 and 0.7. The number ofevents is normalized in both cases to 10.7 hours of observation. The position of theoptical source LSI +61 303 (yellow cross) and the 95% confidence level contours for3EG J0229+6151 and 3EG J0241+6103 (green contours) (Hartman, 1999), are alsoshown. The bottom-right circle shows the size of the point spread function of MAGIC(1 σ radius). No significant excess in the number of γ -ray events is detected aroundperiastron passage, while it shows up clearly (9.4 σ statistical significance) at laterorbital phases, in the location of LS I +61 303. (Albert et al., 2006e).questions and even the a possible vacuum refractive index, that might be induced by quantumgravity, can be probed. The set of extragalactic objects which were observed by the MAGICtelescope comprises known TeV-emitting blazars (AGNs with a jet axis close to the line ofsight) as well as VHE γ -ray candidate sources such as selected high- and low-frequency peakedBL Lacs (HBLs and LBLs). Moreover, other non–blazar objects like the ULIRG Arp 220 havealso been observed, although none of these observations resulted in a positive detection so far(Albert et al., 2007b).In section 4.1 the discoveries of VHE γ -rays from candidate sources will be described, whilein section 4.2 the observation of known TeV blazars is reviewed. γ -Rays from Candidate Sources The selection of candidates for VHE γ -ray emitting sources follows criteria based on the spectralproperties of the considered objects at lower frequencies, see e.g. Albert et al. (2007n). Usingboth Synchrotron Self–Compton (SSC) and hadronic models, the spectral energy distributionof the candidate AGN is extrapolated to MAGIC energies to predict its observability. Thepreferred candidates are usually strong X-ray emitters, but selections based on the optical bandhave also been followed. (Albert et al., 2006f) is the first source discovered by MAGIC and one ofthe most distant VHE γ -ray sources known so far. This HBL, which has a redshift of z = 0 . bservations of VHE γ -Ray Sources with the MAGIC Telescope 7 Energy [GeV] -1
10 1 10 - s - T e V m dF / d E · E -8 -7 -6 EGRETWhippleHEGRACANGAROOCelesteStaceeTibetH.E.S.S.Aharonian et al. (2004)MAGIC
Fig. 4
Spectral energy distribution of the γ -ray emission of the Crab Nebula. The measure-ments below 10 GeV are from the EGRET, themeasurements above are from ground-basedexperiments. Above 400 GeV the MAGICdata are in agreement with measurements ofother IACTs. The dashed line represents amodel prediction by Aharonian et al. (2004a).(Albert et al., 2007k). MeV Energy - s - M e V c m d F / d E E -7 -6 -5 -4 -3 MAGIC 95% U.L. Celeste (De Naurois 2002)Whipple (Lessard 2000)Outer Gap (Hirotani 2007)EGRET (Fierro 1998) and 27 GeV exponential cutoffEGRET (Fierro 1998)
Fig. 5
Upper limits (95% CL.) on the pulsed γ -ray flux from the Crab Pulsar; upper lim-its in bins of energy are given by the bluepoints. The upper limit on the cutoff energy ofthe pulsed emission is indicated by the dashedline. The analysis threshold to derive the up-per limit on the cutoff energy is indicated bythe red arrow. (Albert et al., 2007k). was previously observed by Whipple and HEGRA, but only upper flux limits were determined.MAGIC observed it during 8.2 h in January 2005, obtaining a γ -ray signal of 6.4 σ significancein the 87 to 630 GeV energy range. The differential energy spectrum can be fitted by a simplepower law with a photon index of 3 . ± .
4. No time variability on timescales of days was foundwithin statistical errors.
Fig. 6
Differential energy spectrum of PG1553+113 as derived from the combined2005 and 2006 data. The MAGIC CrabNebula energy spectrum and the H.E.S.S. PG1553+113 energy spectrum (Aharonian et al.,2006b) have been included for comparison.(Albert et al., 2007a).
Energy [GeV] ] - s - d N / d E [ T e V c m E -13 -12 -11 -10 -9 G - E Power law fit: dN/dE = N 0.31 – = 2.65 G / NDF = 7.8 / 8 c H.E.S.S., measuredMAGIC, measuredH.E.S.S., intrinsicMAGIC, intrinsicpower law fitcurved power law fit
PG 1553+113z = 0.42
Fig. 7
Observed differential energy spectrumof PG 1553+113 multiplied by E to repre-sent the spectral energy density by H.E.S.S.and MAGIC (open symbols) and source in-trinsic spectra (full symbols), corrected for theattenuation by the EBL assuming z = 0 . PG 1553+113 is a distant BL Lac of undetermined redshift. It was recently detected bythe H.E.S.S. collaboration (Aharonian et al., 2006b) as well as by the MAGIC collaboration(Albert et al., 2007a). A VHE γ -ray signal has been observed by the MAGIC telescope with anoverall significance of 8.8 σ , showing no significant flux variations on a daily timescale. However,the flux observed in 2005 was significantly higher compared to 2006. The MAGIC measurementsreach to substantially lower energies than the HESS measurements do. The differential energyspectrum between 90 and 500 GeV can be well described by a power law with photon indexof 4 . ± .
3, being steeper than that of any other known BL Lac object, see figure 6. Thisspectrum can be used to derive an upper limit on the source redshift. Assuming an EBL modelby Kneiske et al. (2004) and a limit of α int < − . α int of the intrinsic sourcespectrum (Aharonian et al., 2006c), an upper limit on the source redshift of z < .
78 has beenderived. Mazin & Goebel (2007) find that a redshift above z = 0 .
42 implies a possible break ofthe intrinsic spectrum at about 200 GeV. Assuming that such a break is absent, they obtain amuch stronger upper limit of z < .
42, see figure 7.
Mkn 180 (Albert et al., 2006h) is an HBL ( z = 0 . γ -ray emission from thissource. A total of 12.4 h of data were recorded during eight nights, giving a 5 . σ significancedetection. The integral flux above 200 GeV corresponded to 11% of the Crab Nebula flux, andthe differential energy spectrum could be fitted by a power law (photon index 3 . ± . Bl Lacertae (Albert et al., 2007g) is a low-frequency peaked BL Lac (LBL) object at z = 0 . γ -ray signal wasdiscovered with a 5.1 sigma excess in the 2005 data. Above 200 GeV, an integral flux of 3%of the Crab Nebula flux was measured. The differential energy spectrum between 150 and 900GeV is rather steep, with a photon index of − . ± .
5. For the first time, a clear detection ofVHE γ -ray emission from an LBL object was obtained. During the observation, the light curveshowed no large flux variation. The 2006 data show no significant excess.Recently, also was discovered as a source of VHE γ -rays by the MAGICtelescope Albert et al. (2007m). The sensitivity and lower energy threshold of MAGIC as compared to the former generation of γ -ray telescopes, allows a detailed study of the spectral features and flux variations of knownTeV emitters. Mkn 421 (Albert et al., 2007d) is the closest TeV blazar ( z = 0 . γ -ray telescope (Punch et al., 1992).MAGIC has observed this source between November 2004 and April 2005, obtaining 25.6 hof data. Integral flux variations up to a factor of four are observed between different obser-vation nights, although no significant intra–night variations have been recorded, despite thehigh sensitivity of the MAGIC telescope for this kind of search. This flux variability shows aclear correlation between γ -ray and X-ray fluxes, favoring leptonic emission models. The energyspectrum between 100 GeV and 3 TeV shows a clear curvature. After correcting the measuredspectrum for the effect of γ -attenuation caused by the EBL assuming a model of Primack et al.(2005), there is an indication of an inverse Compton peak around 100 GeV. ( z = 0 . bservations of VHE γ -Ray Sources with the MAGIC Telescope 9 ] - s - F ( - G e V ) [ c m -9 × ] - s - F ( - G e V ) [ c m -9 × ] - s - F ( - G e V ) [ c m -9 × Fig. 8
VHE γ -ray light curve for the night2005 July 9 with a time binning of 4 minutes,and separated in 3 different energy bands, fromthe top to the bottom, 0.25-0.6 TeV, 0.6-1.2TeV, 1.2-10 TeV. The vertical bars denote 1 σ statistical uncertainties. For comparison, theCrab emission is also shown as a lilac dashedhorizontal line. The vertical dot-dashed line di-vides the data into ’stable’ (i.e., pre-burst) and’variable’ (i.e., in-burst) emission. The hori-zontal black dashed line represents the aver-age of the ’stable’ emission. The ’variable’ (in-burst) of all energy ranges were fitted with aflare model. (Albert et al., 2007h). Fig. 9
Overall SED from Mrk501. Opticaldata from the KVA Telescope: green cir-cle, X-ray data from RXTE ASM for June30 (black points), for July 9 (red) and forthe other nights combined (light blue). VHEdata from MAGIC: black points (June 30),red points (July 9), green points (’high flux’data-set), dark blue points (’medium flux’data-set), and light blue points (’low flux’data-set). The VHE spectra are corrected forEBL extinction using (Kneiske et al., 2004)’s’Low’ EBL model. The highest and the loweststate were fitted with a one zone SSC model.(Albert et al., 2007h). for a signal on a 4 sigma level (Aharonian et al., 2004b). MAGIC obtained a VHE γ -ray signalwith 11.0 σ significance from 23.1 h of data (Albert et al., 2007c), measuring its energy spectrumfrom 140 GeV to 5 TeV. The source was in the quiescent state during the observations, with aflux level compatible with the HEGRA results, but showing a softer spectrum. ( z = 0 . γ -rayflare without any counterpart in X-rays (Krawczynski et al., 2004). This behavior cannot beeasily explained by the SSC mechanism in relativistic jets that successfully explains most ofthe VHE γ -ray production in other HBLs. MAGIC observed this object during 6 h in 2004,when it was in low activity both in optical and X-ray bands, detecting a γ -ray signal with8.2 σ significance (Albert et al., 2006c). The differential energy spectrum between 180 GeV and2 TeV can be fitted with a power law of photon index 2 . ± .
14, which is consistent withthe slightly steeper spectrum seen by HEGRA at higher energies (Aharonian et al., 2003), alsoduring periods of low X-ray activity.
Mkn 501 (Albert et al., 2007h) is a close TeV blazar ( z = 0 .
34 h in moderate moonlight conditions. The source was in the low state (30-50% of the CrabNebula flux for
E >
200 GeV) during most of the observation time, but showed two episodes offast and intense flux variability, with doubling times of about 2 minutes, see figure 8. The energyspectrum was measured from 100 GeV up to 5 TeV. Changes in the spectral slope with the fluxlevel have been observed for the first time on timescales of about 10 minutes. Figure 9 showsthe overall SED of Mrk501 for different days as well as ’high’, ’medium’ and ’low’ flux datasets. Recently, the timing of photons observed by the MAGIC gamma-ray telescope during thisflare was used to probe a vacuum refractive index, that might be induced by quantum gravity(Albert et al., 2007p).
A wealth of gamma ray bursts (GRBs) have been observed since their first detection more than30 years ago. Nevertheless, the processes leading to these extraordinarily energetic outburstsare still largely unknown. EGRET, a satellite γ -ray telescope, measured γ -rays up to 18 GeV(Hurley et al., 1994), and measuring them at higher energies with the MAGIC telescope wouldsubstantially help to understand them better. Such observations would also give a clue as totheir distance, because high-energy γ -rays at larger distances are increasingly strongly absorbedby the intergalactic infrared background radiation. More precise localization, as possible inMAGIC, would also help in identifying them with known sources. Possible correlations betweenflux variations at different energies might even allow setting limits for the invariance of thespeed of light, which some models of quantum gravity claim to be violated.MAGIC being built specifically also for GRB observations, by minimizing weight and in-stalling powerful driving motors, an alert for GRB (as they come from satellite experiments)triggers the re-orientation of the telescope and the start of the new observation within a max-imum time (depending on the position in the sky) of 40 seconds. On 13 July 2005, triggeredby such an alarm from Swift-BAT, MAGIC succeeded for the first time to track a GRB dur-ing its prompt phase (Albert et al., 2006d). No significant radiation at high energy was seen.Thereafter, nine additional GRBs have been targeted with the MAGIC telescope, but in neitherdata set any evidence for a γ -ray signal was found. Upper limits for the flux were derived forall events. For the bursts with measured redshift, the upper limits are compatible with a powerlaw extrapolation, when the intrinsic fluxes are evaluated taking into account the attenuationdue to the scattering in the Metagalactic Radiation Field (MRF) (Albert et al., 2006j). We know today, from measuring gravitational effects, that the visible Universe represents onlya fraction of the matter in the Universe: some 75% of all matter cannot be seen, and is called”dark matter”. Dark matter cannot be made up of the same constituents as visible matter,and must be ”non-baryonic”. Some theories predict that, with very low probability, such non-baryonic particles upon collision can produce VHE γ -rays. Observing such annihilation productswould, of course, be an epochal discovery for cosmology and astrophysics. MAGIC dedicatessome observation time to such searches, despite the small probabilities involved (Bartko et al.,2005b). Using the MAGIC telescope, seven galactic and nine extra-galactic sources of VHE γ -radiationhave been observed. Nine of these objects have been detected before in VHE γ -rays. The highsensitivity and the low energy threshold of the MAGIC telescope allowed detailed studies of bservations of VHE γ -Ray Sources with the MAGIC Telescope 11 the spectral features of these sources, as well as the observation of flux variability on shorttimescales.The MAGIC collaboration is currently constructing a second telescope on the same site atthe Roque de los Muchachos Observatory, which will operate in stereo mode with the MAGICtelescope improving the overall sensitivity (Teshima et al., 2005). The MAGIC II telescope isa clone of the existing one with one main improvement: a fine pixelized camera with a clusterdesign that will allow an upgrade of the photomultipliers to hybrid photon detectors once thistechnology is ready to be used. The estimated sensitivity of a system of two MAGIC telescopesis a factor of two better than the present sensitivity of the MAGIC telescope.Acknowledgements. We thank the IAC for the excellent working conditions at the ORM in LaPalma. The support of the German BMBF and MPG, the Italian INFN, the Spanish CICYTis gratefully acknowledged. This work was also supported by ETH research grant TH-34/04-3,and the Polish MNiI grant 1P03D01028. References
Aharonian, F. et al. (HEGRA Collab.), 2003, A&A, 406, L9.Aharonian, et al. (HEGRA Collab.), 2004, ApJ, 614, 897.Aharonian, et al. (HEGRA Collab.), 2004, A&A, 421, 529.Aharonian, F. et al. (HESS Collab.), 2005, Science, 309, 746.Aharonian, F. et al. (HESS Collab.), 2006a, ApJ, 636, 777.Aharonian, F. et al. (HESS Collab.), 2006b, A&A, L19.Aharonian, F. et al. (HESS Collab.), 2006c, Nature, 440, 1018.Aharonian, F. et al. (HESS Collab.), 2006d, PRL, 97, 249901.Albert, J. et al. (MAGIC Collab.), 2006a, ApJ, 637, L41.Albert, J. et al. (MAGIC Collab.), 2006b, ApJ, 638, L101.Albert, J. et al. (MAGIC Collab.), 2006c, ApJ, 639, 761.Albert, J. et al. (MAGIC Collab.), 2006d, ApJ, 641, L9.Albert, J. et al. (MAGIC Collab.), 2006e, Science 312, 1771.Albert, J. et al. (MAGIC Collab.), 2006f, ApJ, 642, L119.Albert, J. et al. (MAGIC Collab.), 2006g, ApJ, 643, L53.Albert, J. et al. (MAGIC Collab.), 2006h, ApJ, 648, L105.Albert, J. et al. (MAGIC Collab.), 2006i, NIM submitted, astro-ph/0612385.Albert, J. et al. (MAGIC Collab.), 2006j, ApJ in press, astro-ph/0612548.Albert, J. et al. (MAGIC Collab.), 2007a, ApJ, 654, L119.Albert, J. et al. (MAGIC Collab.), 2007b, ApJ, 658, 245.Albert, J. et al. (MAGIC Collab.), 2007c, ApJ, 662, 892.Albert, J. et al. (MAGIC Collab.), 2007d, ApJ 663, 125.Albert, J. et al. (MAGIC Collab.), 2007e, ApJ, 664, L87.Albert, J. et al. (MAGIC Collab.), 2007f, ApJ, 665, L51.Albert, J. et al. (MAGIC Collab.), 2007g, ApJ, 666, L17.Albert, J. et al. (MAGIC Collab.), 2007h, ApJ submitted, astro-ph/0702008.Albert, J. et al. (MAGIC Collab.), 2007i, ApJ submitted, astro-ph/0702077.Albert, J. et al. (MAGIC Collab.), 2007j, Astropart. Phys. submitted, astro-ph/0702475.Albert, J. et al. (MAGIC Collab.), 2007k, ApJ submitted, arXiv:0705.3244.Albert, J. et al. (MAGIC Collab.), 2007l, A&A submitted, arXiv:0706.4065.Albert, J. et al. (MAGIC Collab.), 2007m, ApJL submitted, arXiv:0706.4435.Albert, J. et al. (MAGIC Collab.), 2007n, ApJ submitted, arXiv:0706.4453.2 H. Bartko for the MAGIC collaboration