Draftversion June22,2014 PreprinttypesetusingLATEXstyleemulateapjv.5/2/11 YOUNG AGN OUTBURST RUNNING OVER OLDER X-RAY CAVITIES A´kos Bogda´n1,6, Reinout J. van Weeren1,6, Ralph P. Kraft1, William R. Forman1, Scott Randall1, Simona Giacintucci2,3, Eugene Churazov4, Christopher P. O’Dea5, Stefi A. Baum5, Jacob Noell-Storr5, and Christine Jones1 1SmithsonianAstrophysicalObservatory,60GardenStreet,Cambridge,MA02138, USA;[email protected] 2DepartmentofAstronomy,UniversityofMaryland,CollegePark,MD20742, USA 3JointSpace-Science Institute, UniversityofMaryland,CollegePark,MD20742-2421, USA 4Max-Planck-Institutfu¨rAstrophysik,Karl-Schwarzschild-str. 1,85741GarchingbeiMu¨nchen, Germanyand 4 5RochesterInstitute ofTechnology, 84LombMemorialDrive,Rochester,NY14623, USA 1 Draft version June 22, 2014 0 2 ABSTRACT n Although the energetic feedback from active galactic nuclei (AGN) is believed to have a profound a effect on the evolutionof galaxiesand clusters of galaxies,details of the AGN heating remainelusive. J Here,westudyNGC193–anearbylenticulargalaxy–basedonX-ray(Chandra)andradio(VLAand 3 GMRT) observations. These data revealthe complex AGN outburst history of the galaxy: we detect 2 a pair of inner X-ray cavities, an outer X-ray cavity, a shock front, and radio lobes extending beyond the inner cavities. We suggestthat the inner cavities were produced 78 Myr ago by a weakerAGN ] ∼ outburst, while the outer cavity, the radio lobes, and the shock front are due to a younger (13 26 A − Myr) and (4 8) times more powerful outburst. Combining this with the observed morphology of G − NGC 193, we conclude that NGC 193 likely represents the first example of a second, more powerful, . AGN outburst overrunning an older, weaker outburst. These results help to understand how the h outburst energy is dissipated uniformly in the core of galaxies, and therefore may play a crucial role p in resolving how AGN outbursts suppress the formation of large cooling flows at cluster centers. - o r Subject headings: galaxies: individual (NGC 193) — galaxies: active — radio continuum: galaxies — t s X-rays: galaxies — X-rays: ISM a [ 1 1. INTRODUCTION 2004; Vernaleo & Reynolds 2006). While these stud- v There is a wide agreementthat the energetic feedback ies addressed various aspects of the interaction, it re- 6 from AGN plays a pivotalrole in the evolution of galax- mains unclear if the models reflect the properties of real 7 ies. Powerful AGN outbursts are capable of quenching AGN feedback. To better understand the interaction 1 between AGN feedback and the surrounding interstel- the active star formation in massive galaxies, thereby 6 lar/intracluster material, it is crucial to explore nearby producing the population of “red and dead” ellipticals. . systems that exhibit signatures of past outbursts. 1 The energy input from AGN can reheat the cooling In this paper, we study the lenticular galaxy NGC 0 gas, and hence suppress star formation and maintain 193 (aka UGC 408), which hosts a luminous FR-I ra- 4 thepassivenatureofellipticals(e.g.Springel et al.2005; dio source, 4C+03.01. The galaxy is located in a poor 1 Croton et al. 2006; McNamara & Nulsen 2007). More- : over,the scalingrelations betweengalaxiesandtheir su- group, where the most optically luminous elliptical is v NGC 194 (Geller et al. 1983). For the distance of NGC permassive black holes (BHs) may also be explained by Xi AGN feedback (e.g. Silk & Rees 1998; Di Matteo et al. 193 we adopt 63 Mpc (1′ = 18.3 kpc). Hubble Space Telescope imagesofNGC193showprominentdustlanes ar 20I0n5).galaxies, galaxy groups, and galaxy clusters (Noel-Storr et al. 2003), and low level (< 0.1 M⊙ yr−1) star formation. The Galactic column density towards the commonly observed signatures of powerful AGN NGC 193 is 2.6 1020 cm−2 (Kalberla et al. 2005). outbursts are cavities in the X-ray gas distribu- × NGC 193 has a complex structure with several phe- tion (e.g. Forman et al. 2005; McNamara et al. 2009; nomena detected, such as X-ray cavities, surface bright- Bogd´an et al.2011). JetsemanatingfromtheBHinflate nessedges,radiolobes,andparsec-scaleradiojets. While radio lobes, which displace the surrounding hot X-ray similar features were studied separately in other galax- emitting gas, possibly drive shocks, and create bubbles ies, NGC 193 uniquely comprises all of these. Given the of relativistic plasma. In a handful of systems multiple complexity of NGC 193, we present our results in two X-ray cavities are detected originating from consecutive papers. Here, we provide evidence that a young, more AGN outbursts (e.g. Randall et al. 2011; Blanton et al. powerful, AGN outburst overran an older, weaker out- 2011),whichallowsanexplorationofAGNoutbursthis- burst. In the followingpaper, we will explore the overall tory. morphologyof NGC 193,study its large-scaledynamics, Given the importance of AGN feedback, numerical and investigate its central regions. studies have attempted to describe the interaction of AGN with their surrounding gas (e.g. Ruszkowski et al. 6EinsteinFellow 2 BOGDA´N ET AL. TABLE 1 The list of analyzedChandra observations. Galaxy ObsID Center coordinates Tobs (ks) Tfilt (ks) Instrument Obs. date NGC193 4053 00h39m18.60s+03d19’52” 29.1 14.5 ACIS-S 2003Sep01 NGC193 11389 00h39m18.60s+03d19’53” 93.9 93.9 ACIS-S 2009Aug21 Fig.1.—Topleft: Chandra 0.3−2keVbandimageofa6′×6′ regionofNGC193. Backgroundissubtractedandvignettingcorrection isapplied. Pointsourcesareexcludedandtheirlocationisfilledwiththelocalbackgroundvalue. TheimageissmoothedwithaGaussian kernel withawidth of2′′. Top right: 1.4GHzVLAimage ofthe sameregion, showingthe slightlyasymmetric radiolobes. Bottom left: SameChandra imageasinthetopleftpanel,butoverlaidaretheintensitylevelsofthe610MHzGMRT image. Bottom right: Residual imageofNGC193inacircularregionwith150′′ radius. Theshockfrontextendstothesouthandpetersattheeast–adetaileddiscussion willbepresentedinanupcomingpaper. AGN OUTBURST RUNNING OVER X-RAY CAVITIES 3 2. DATAREDUCTION In a simple picture, the disturbed X-ray gas distribu- tion,thesurfacebrightnessedges,andtheradiolobesare 2.1. X-ray data dueto arecentAGNoutburst. Inthis section,we estab- NGC 193 was observed by Chandra for a total of 123 lish the basic properties of this outburst, specifically the ks. The data were reduced with standard CIAO1 soft- expansion velocity of the lobes and the occurrence time ware package tools (CIAO version4.5; CALDB version of the outburst. 4.5.6). To characterize the X-ray surface brightness edge at The data analysis was performed following 2′ central distance, we build surface brightness and Bogd´an & Gilfanov (2008). First, we filtered the ∼temperatureprofilesinthe 0.3 2keVbandusing circu- flare contaminated time intervals, which resulted in a lar wedges along the eastern a−nd western sides of NGC total net exposure time of 108 ks. Bright point sources 193 with an opening angle of 60◦ (Figure 2). The pro- were identified and masked out. We subtracted the files show a sharp surface brightness edge and a tem- instrumental and sky background components by using perature jump at the east of NGC 193, which suggest the ACIS “blank-sky” files. To correct for temporal that this is a shock front produced by the supersonic variations in the normalization of the background expansionof the radio lobes – for the 610 MHz radio in- components, the “blank-sky” files were renormalized by tensitylevelsoverlaidontheX-rayimageseethebottom the ratio of the count rates in the 10 12 keV band. left panel of Figure 1. To estimate the expansion veloc- − We built exposure maps – assuming a thermal model ity,weusetheRankine-Hugoniotjumpconditionsacross with kT = 0.9 keV, 0.3 Solar metallicities, and Galactic the shock front and the gas densities/temperatures up- column density – to account for the vignetting effects. stream and downstream (Markevitch & Vikhlinin 2007; Russell et al. 2010). From the density drop of n /n = 2.2. Radio data 1.60 0.15,weobtainaMachnumber ofM =1.421+01.12, NGC 193 was observed with the Very Large Array ± −0.10 while from the temperature drop of T /T =1.40 0.09 (VLA) at 1.4 GHz and 4.9 GHz between March 2007 we deduce M = 1.13+0.12. The lower2Ma1ch numb±er ob- andMay2008(Laing et al.2011). Theinitialdatareduc- −0.13 tained from the temperature jump is presumably due to tionstepswerecarriedoutwiththeNRAOAstronomical projection effects. To account for this effect, deprojec- Image Processing System package. Gain solutions were tion could be employed, which would result in a some- obtained for the calibrator sources and transferred to whathigher Machnumber. However,we do notperform NGC 193. Theflux-scalewassetbythe Perley & Taylor a deprojection analysis, since the hot gas distribution (1991) extension to the Baars et al. (1977) scale. Each exhibits significant deviations from spherical symmetry, dataset from a specific VLA configuration was then im- which would result in large systematic uncertainties in agedandseveralroundsofphaseselfcalibrationwerecar- the deprojected gas parameters. ried out to refine the calibration. A final step of phase To estimate the age of the AGN outburst, we assume and amplitude self calibration was performed. We im- ported the data in CASA2 and combined the data from that the radio lobes are on the plane of the sky (θ = 90◦), they expandwith aconstantvelocityofM =1.0 allconfigurations. Wethenimagedthetwodatasetswith 1.5 (480 720 km s−1 in a 0.9 keV plasma), and the−y the multi-scale clean algorithm. To make spectral index extend to−112′′ ( 34.2 kpc). Thus, the age of the radio maps,wecreatedimageswithuniformweighing,tocom- ∼ lobes is 47 70 Myr. This estimate may be modified by pensateforthedifferentsamplingintheuv-lane. Wealso − two factors. First, the expansion may have been faster employedGaussianuv-taperstomatchtheresolutionsof in the past (Churazov et al. 2000), which could decrease both datasets. The Giant Metrewave Radio Telescope (GMRT) ob- the expansion time by a factor of 3/5 to 28 42 Myr. − Second, if the axis of the radio lobes does not lie in the served NGC 193 at 235 MHz and 610 MHz in plane of the sky (i.e. θ = 90◦), as suggested by the August 2008 and August 2007, respectively. The 6 GMRT data presented in this work come directly from asymmetric radio lobes, then their real extent is larger than the projected one. Therefore, the expansion time Giacintucci et al. (2011). that takes into account these effects is (28 42)/sinθ 3. RESULTS Myr. − 3.1. A simple picture of NGC 193 Another age estimate on the radio lobes can be ob- tained from computing their radiative age. To this end, In Figure 1 (top left panel) we show the 0.3 2 keV − we extracted the flux densities at 1.4 GHz and 4.9 GHz bandChandra imageofNGC193,whichrevealsthepres- across the lobes and computed the corresponding spec- ence of diffuse X-ray emission originating from hot ion- tral index (α). As shown in the spectral index map izedgas. Thegasdistributionisdisturbed,andshowsthe (Fig. 3), α remains fairly flat along the east-west axis presence of cavities and sharp surface brightness edges. and steepens along the lobes. Assuming that the last The 1.4 GHz VLA image (Figure 1 top right panel) re- electron acceleration occurred at the outer edge of the veals somewhat asymmetric bipolar radio lobes, extend- lobes, we fitted the observed steepening using a Jaffe- binigpotloar∼ra2d′ioinlotbheeseaarset-nwoetsotndlyiredcettieocnt.edInatter1e.s4tiGnHglzy,,bthuet Perola model assuming that νbreak ∝ d−2, where d is thedistancefromtheinjectionsource. Usingthebest-fit alsointhe4.9GHzVLAandthe235MHzand610MHz ν = 2.0 1.2 GHz and the “revised” equipartition GMRT images. These evidencesindicate that the BHof break ± magnetic field of 16 µG obtained from Beck & Krause NGC 193 was active in the recent past. (2005), we estimate that the age of the lobes is 13 26 − 1 http://cxc.harvard.edu/ciao/ Myr. This estimate should be considered as an upper 2 http://casa.nrao.edu/ limit, since we assumed that the magnetic field is con- 4 BOGDA´N ET AL. 1.5 East East 2m] West 1.4 West 2c/c10-17 1.3 e s s/arc keV] 1.2 s [erg/ ature [ 1.1 brightnes10-18 Temper 0. 91 e 0.8 c a urf 0.7 S 10-19 0.6 50 100 150 200 250 0 50 100 150 200 250 Radius [arcsec] Radius [arcsec] Fig. 2.— Surface brightness (left panel) and temperature (right) profiles of NGC 193 extracted from circular wedges with an opening angleof60◦towardthewestandeast. Thethick(red)verticallineindicatesthepositionofthesurfacebrightnessjump,whichiscoincident withtheextent oftheradiolobetowardstheeast. Thethin(blue)verticallineshowstheextent oftheradiolobetowardsthewest. NGC193,(iii)brightrims surroundingtheinnercavities, and (iv) sharp surface brightness edges. While the outer cavity and the surface brightness edge is in good agree- mentwiththedistributionoftheeasternradiolobe(Fig- ure1bottomleft),theinnercavitypairisorientedinthe north-south direction as opposed to the east-west align- mentoftheradiolobes. Additionally,the radioemission is not enhanced at the inner cavity, and the radio lobes extendsignificantlybeyondthem. Thissuggeststhatthe inner and outer cavities may originate from two distinct AGN outbursts. To estimate the occurrence time of the outburst that inflated the inner cavities, we derive their buoyant rise time following McNamara & Nulsen (2007). Assuming that the cavities have a radius of 37.5′′ and using the gas temperature of 0.9 keV we find that t = 78 Myr. b Weemphasizethatduetopossibleprojectioneffectsthis valueshouldbeconsideredasalowerlimit. The absence Fig.3.—Spectralindexmapofa6′×6′regionaroundNGC193 of radio emission directly associated with the inner cav- inferredfromthe ratio of the 1.4GHz and 4.9 GHz VLA images. Overplotted are the 4.9 GHz VLA contours, where the contour ities also suggests an old outburst. Using the 235 MHz levelsarespacedat[1,2,4,8,...]×3σrms. GMRT image and assuming a magnetic field of 10 µG (Parma et al. 2007), we estimate that the fading time of theradioemissionis 50Myr,whichplacesalowerlimit stant and uniform across the lobe and expansion energy ∼ on the age of the outburst. Note that the precise value losses are negligible. At the face value, the two different of the magnetic field is not known, yielding a somewhat ageestimates arebroadlyconsistent,albeitthe radiative uncertain estimate. age may be somewhat shorter. These results hint that the morphological features of NGC193originatepresumablyfromtwoconsecutiveout- 3.2. Two consecutive AGN outbursts bursts. In this picture, an old outburst ( 78 Myr ago) Although previously we considered a simple scenario, inflatedtheinnerX-raycavitiesthatareb∼uoyantlygrow- inwhichtheobservedX-rayandradiofeaturesaredueto ing at the present epoch. This outburst was followed a single AGN outburst, this picture cannot give a com- by ayoungeroutburst,that producedthe supersonically plete descriptionofNGC 193fortworeasons. First, this expanding outer cavity, the shock front, and the radio scenariocannotaccountfor all majormorphologicalfea- lobes. A consequence of this interpretation is that the tures. Second,thispictureresultsintimescaleestimates inner cavities were inflated before the outer cavities, and that are in tension with each other. hence the presently observed younger radio jets overran Toexploreandbetteremphasizethemorphologyofthe the older X-ray cavity. X-ray gas, we construct a residual X-ray image. There- The scenario, in which the younger outburst overran fore,wesubtractanazimuthallyaveragedsurfacebright- an older outburst, can resolve the mismatch between ness image from the 0.3 2 keV band Chandra image. the estimated time scales of the younger AGN outburst − The thus obtained image (Figure 1 bottom right) shows (Section 3.1). Specifically, the expansion time scale of the following features: (i) a pair of inner cavities in the (28 42)/sinθMyrobtainedfromtheRankine-Hugoniot centralregions,(ii) anouter cavity in the easternside of − AGN OUTBURST RUNNING OVER X-RAY CAVITIES 5 jump conditions is likely overestimated. Indeed, a high how do AGN outbursts suppress the formation of large momentum jet can cross the low density region nearly cooling flows at cluster centers. It is clear that regu- with the speed of light, which implies that the propaga- lar AGN outbursts can provide sufficient energy to off- tion time within the X-ray cavities is virtually negligi- set radiative losses and prevent the formation of cooling ble. Assuming that the cavities extend to 50′′, the radio flows. How the energy of the outburst is dissipated uni- lobes expand at the estimated M = 1.0 1.5 velocity formly in the core and not at larger radii is uncertain in the 50 112′′ region. Taking into acco−unt the vari- (O’Neill & Jones 2010). Buoyantly rising bubbles will − ouscorrections(Section3.1),theageofAGNoutburstis dissipatetheirenergynon-isentropically,butmuchofthis (16 23)/sinθ Myr,whichisconsistentwiththe 13 26 will occur outside the region of largest radiative losses − − Myr age estimate obtained from the radio data. (Vernaleo & Reynolds 2006). Likewise, regular low-level By measuringthe cavity power(P ) we estimate the AGN outbursts will generate sounds waves (Begelman cav energy injected to the X-ray emitting gas by the two 2001; Churazov et al. 2002), but efficient heating of the AGN outbursts. We derive P from the minimal en- gas in the core requires that the viscosity of the gas is cav ergy required to inflate the cavity and the age of the a significant fraction of the Spitzer value. The entropy cavity (McNamara & Nulsen 2007; Bogd´an et al. 2011). increasedue to weakshocks is the mostlikely method of For relativistic plasma, the minimal energy is obtained increasing the entropy of the gas in the core and offset- from the cavity enthalpy, H =4pV, where p is the aver- tingtheradiativelosses(McNamara & Nulsen2007),but age pressure and V is the volume of the cavity. how this energy input is uniformly distributed through- Wedescribetheinnercavitieswithtwocircularregions out the core remains unclear. with 37.5′′ radius (Figure 1 bottom right) and assume If a strong AGN outburst is driven through an older, spherical symmetry. Using the best-fit spectrum, we de- presently buoyant, outburst, the shock from the current rivetheaverageelectrondensityofne =4.2 10−3cm−3, outburstwillberapidlydriventhroughtheoldradiobub- × and the average pressure as p = 1.9nekT = 1.0 blessincethe soundspeedis high(muchhigherthanthe 10−11 erg cm−3. Thus, the total AGN work done b×y shockspeedinthethermalgas,ashighasc/√3iftheold the cavities is 1.5 1058 erg. Assuming that the age radio bubble is relativistic plasma). Thus, the old radio × of the inner cavity is 78 Myr, the total cavity power is bubble will effectively “isotropize” the shock from the Pcav =6.1 1042 erg s−1. more recent outburst, more uniformly distributing the × The outer cavity is described with a circular region heating throughout the core. Additionally, if the older with a circular region with 40′′ radius (Figure 1 bot- radiobubbleispartiallyshredded,thepassingshockwill tom right). Since the outer cavity is not detected on turnthesmallerclumpsofradioplasmaintovortexrings, the western side of NGC 193, we base our calculations andthe vorticityinthe gascansignificantlyincreasethe on one cavity and assume spherical symmetry. From efficiency of heating the ICM (Heinz & Churazov 2005). the best-fit spectrum we obtain the average density of ThecombinationofvariablepowerAGNjetswiththedy- ne = 3.5 10−3 cm−3 and the average pressure of namic ICM (Heinz et al. 2006) are likely the key to bal- p = 1.9nek×T = 1.1 10−11 erg cm−3, and hence com- ancingtheradiativelossesinclustercoresandpreventing pute the cavity ene×rgy of 1.0 1058 erg. Since this the formationof large cooling flows. A more careful and × value refers to one cavity, the total cavity energy is systematicexaminationofarchivalChandra observations 2.0 1058 erg. Assuming the cavity age of 13 26 of groups and clusters would provide stronger limits on × − Myr, the total cavity power of the young outburst is the frequency with which AGN outbursts overlap. Pcav = (2.4 4.9) 1043 erg s−1. Thus, the more re- The observationaldata points out that the inner cavi- − × cent outburst, that overran the inner X-ray cavities, is ties that originate fromthe older outburst have a north- (4 8) times more powerful than the older outburst. south orientation,while the younger radio lobes and the − outercavityarepositionedalongtheeast-westdirection. 4. DISCUSSION It is feasible that the jet axis changed by 90◦ between Our observation of two AGN outbursts in NGC 193 the two AGN outbursts, which may have∼been triggered is the first example of a second, more powerful, AGN byarecentmergerthatchangedtheblackholespindirec- outburst overrunning an older, weaker outburst. Clear tion,andhence the orientationofthe jet (Nixon & King examples of multiple outbursts in group environments, 2013; O’Dea et al. 2013). However, the observed mor- such as NGC 5813 (e.g. Randall et al. 2011) have been phology can also be obtained without the reorientation reported, but all these cases the outbursts are clearly ofjets. AccordingtothescenariooutlinedinSection3.2, spatially separated, and the oldest outburst is also the the inner cavities are due to a relatively weak outburst. mostpowerful. This is almostcertainly anobservational Following a weak and short outburst with an arbitrary bias since it is easiest to distinguish the outbursts when jet orientation, the inner cavities could have expanded they are well separated, and the oldest outburst is only and moved to follow the gravitational potential or even detectable if it is the strongest since the shock gener- just the motion of gas in the core. Since the galaxy is ated by the oldest outburst will also have weakened the approximately elongated in the northeast-southwest di- most. A weak pressure wave (i.e. transsonic/subsonic) rections,the buoyantlymovinglobeswouldmostlymove from an old outburst is likely undetectable in any but in the same direction. The younger and more powerful the brightest, best-observed systems. jet was presumably not affected by buoyant forces, and Given the small number statistics and difficulty in henceitcouldpropagateintheeast-westdirection,which identifying outbursts that have overrun each other, it would make it appear that the new and old jet axis are is not clear how common they are. Multiple, interacting different. Thus,while a changein the jet axis is feasible, outbursts could be crucial in resolving one of the long it is not required to explain the observational data. standing issues in galaxy group/galaxycluster physics – 6 BOGDA´N ET AL. Acknowledgements. We thank Paul Nulsen for helpful dis- AssociatedUniversities,Inc. A´BandRvWacknowledgesup- cussions. This research made use of Chandra data provided port provided by NASA through Einstein Postdoctoral Fel- by the CXC. The VLA is a facility of NRAO, which is a fa- lowship awarded bytheCXC,which is operated bytheSAO cility of the NSF operated under cooperative agreement by for NASAundercontract NAS8-03060. REFERENCES Baars,J.W.M.,Genzel,R.,Pauliny-Toth,I.I.K.&Witzel,A., Laing,R.A.,Guidetti,D.,Bridle,A.H.,etal.,2011,MNRAS, 1977, A&A,61,99 417,2789 Baldi,A.,Forman,W.,Jones,C.,etal.,2009, ApJ,707,1034 Markevitch,M.&Vikhlinin,A.,2007,PhR,443,1 Beck,R.&Krause,M.,2005,AN,326,414 McNamara,B.R.&Nulsen,P.E.J.,2007,ARA&A,45,117 Begelman,M.C.,2001, ASPC,240,363 McNamara,B.R.,Kazemzadeh, F.,Rafferty, D.A.,etal.,2009, Blanton,E.L.,Randall,S.W.,Clarke,T.E.,etal.,2011,ApJ, ApJ,698,594 737,99 Nixon,C.&King,A.,2013,ApJ,765,L7 Bogd´an,A´.&Gilfanov,M.,2008, MNRAS,388,56 Noel-Storr,J.,Baum,S.,VerdoesKleijn,G.,etal.,2003,ApJS, Bogd´an,A´.,Kraft,R.P.,Forman,W.R.,etal.,2011,ApJ,743, 148,419 59 O’Dea,C.P.,Baum,S.A.,Tremblay,G.R.,etal.,2013, ApJ, Churazov,E.,Forman,W.,Jones,C.&B¨ohringer,H.,2000, 771,38 A&A,356,788 O’Neill,S.M.&Jones,T.W.,2010,ApJ,710,180 Churazov,E.,Sunyaev, R.,Forman,W.,B¨ohringer,H.,2002, Parma,P.,Murgia,M.,deRuiter,H.R.,etal.,2007, A&A,470, MNRAS,332,729 875 Croton,D.J.,Springel,V.,White,S.D.M.,etal.,2006, Perley,R.A.&Taylor,G.B.,1991,AJ,101,1623 MNRAS,365,11 Randall,S.W.;Forman,W.R.;Giacintucci,S.,etal.,2011, ApJ, DiMatteo,T.,Springel,V.&Hernquist,L.,2005,Nature,433, 726,86 604 Ruszkowski,M.,Bru¨ggen,M.&Begelman,M.C.,2004,ApJ, Forman,W.,Nulsen,P.,Heinz,S.etal.,2005,ApJ,635,894 611,158 Geller,M.J.&Huchra,J.P.,1983,ApJS,52,61 Russell,H.R.,Sanders,J.S.,Fabian,A.C.,etal.,2010, Giacintucci,S.,O’Sullivan,E.,Vrtilek,J.,etal.,2011, ApJ,732, MNRAS,406,1721 95 Silk,J.&Rees,M.J.,1998, A&A,331,L1 Heinz,S.&Churazov,E.,2005,ApJ,634,L141 Springel,V.,DiMatteo, T.&Hernquist,L.,2005,ApJ,620,L79 Heinz,S.,Bru¨ggen,M.,Young,A.&Levesque, E.,2006, Vernaleo,J.C.&Reynolds,C.S.,2006,ApJ,645,83 MNRAS,373,L65 Kalberla,P.M.W.,Burton,W.B.,Hartmann,D.,etal.,2005, A&A,440,775