Fossil group origins - VI. Global X-ray scaling relations of fossil galaxy clusters
A. Kundert, F. Gastaldello, E. D'Onghia, M. Girardi, J. A. L. Aguerri, R. Barrena, E. M. Corsini, S. De Grandi, E. Jiménez-Bailón, M. Lozada-Muñoz, J. Méndez-Abreu, R. Sánchez-Janssen, E. Wilcots, S. Zarattini
aa r X i v : . [ a s t r o - ph . C O ] S e p Mon. Not. R. Astron. Soc. , 1–17 (2015) Printed 20 July 2018 (MN L A TEX style file v2.2)
Fossil group origins – VI. Global X-ray scaling relations offossil galaxy clusters
A. Kundert ⋆ , F. Gastaldello , E. D’Onghia † , M. Girardi , , J. A. L. Aguerri , ,R. Barrena , , E. M. Corsini , , S. De Grandi , E. Jiménez-Bailón ,M. Lozada-Muñoz , J. Méndez-Abreu , R. Sánchez-Janssen , E. Wilcots ,and S. Zarattini , , Department of Astronomy, University of Wisconsin-Madison, 475 N. Charter St., Madison, WI 53706, USA INAF – IASF Milano, via E. Bassini 15, I-20133 Milano, Italy Dipartimento di Fisica-Sezione Astronomia, Università degli Studi di Trieste, via Tiepolo 11, I-34143, Trieste, Italy INAF – Osservatorio Astronomico di Trieste, via Tiepolo 11, I-34143, Trieste, Italy Instituto de Astrofísica de Canarias, C/ Vía Láctea s/n, E-38200 La Laguna, Tenerife, Spain Departamento de Astrofísica, Universidad de La Laguna, E-38205 La Laguna, Tenerife, Spain Dipartimento di Fisica e Astronomia ‘G. Galilei’, Università di Padova, vicolo dell’Osservatorio 3, I-35122 Padova, Italy INAF – Osservatorio Astronomico di Padova, vicolo dell’Osservatorio 5, I-35122 Padova, Italy INAF – Osservatorio Astronomico di Brera, via E. Bianchi 46, I-23807 Merate, Italy Instituto de Astronomía, Universidad Nacional Autónoma de México, Apartado Postal 70-264, 04510-México DF, Mexico School of Physics and Astronomy, University of St Andrews, North Haugh, St Andrews, KY16 9SS, UK NRC Herzberg Institute of Astrophysics, 5071 West Saanich Road, Victoria, BC, V9E 2E7, Canada
Accepted 2015 August 12. Received 2015 July 15; in original form 2015 March 16
ABSTRACT
We present the first pointed X-ray observations of 10 candidate fossil galaxy groupsand clusters. With these
Suzaku observations, we determine global temperatures andbolometric X-ray luminosities of the intracluster medium (ICM) out to r for sixsystems in our sample. The remaining four systems show signs of significant contam-ination from non-ICM sources. For the six objects with successfully determined r properties, we measure global temperatures in the range . T X . , bolomet-ric X-ray luminosities of . × L X , bol . × erg s − , and estimate masses,as derived from T X , of M & M ⊙ . Fossil cluster scaling relations are constructedfor a sample that combines our Suzaku observed fossils with fossils in the literature.Using measurements of global X-ray luminosity, temperature, optical luminosity, andvelocity dispersion, scaling relations for the fossil sample are then compared with acontrol sample of non-fossil systems. We find the fits of our fossil cluster scaling re-lations are consistent with the relations for normal groups and clusters, indicatingfossil clusters have global ICM X-ray properties similar to those of comparable massnon-fossil systems.
Key words:
X-rays: galaxies: clusters - galaxies: clusters: general - galaxies: groups:general
Fossil galaxy systems are group and cluster mass objectscharacterized by extended, relaxed X-ray isophotes and anextreme magnitude gap in the bright end of the optical lu-minosity function of their member galaxies. Typically, fos- ⋆ E-mail: [email protected] † Alfred P. Sloan Fellow sils are identified with the criteria of a halo luminosity of L X , bol > . × erg s − and a first ranked galaxy morethan 2 R -band magnitudes brighter than the second bright-est galaxy within half the virial radius (Jones et al. 2003).Fossil systems comprise 8-20 per cent of groups and clus-ters in the same X-ray luminosity regime (Jones et al. 2003),and thus determining the origin of the features characteriz-ing these systems is important for understanding the nature c (cid:13) A. Kundert et al. and evolution of a significant fraction galaxy groups andclusters.The features of fossil systems seem to fulfil theoreti-cal predictions that the Milky Way luminosity ( L *) galax-ies in a group will merge into a central bright elliptical inless than a Hubble time, but the time-scale for the cool-ing and collapse of the hot gas halo is longer (Barnes 1989;Ponman & Bertram 1993). Indeed the first fossil group dis-covered, RX J1340.6+4018 (Ponman et al. 1994), appearedas a solitary bright elliptical located in the centre of a group-sized X-ray luminous halo. It was thought the central galaxyof this group was the final merger remnant of the formergroup galaxies, and hence this object was named a ‘fos-sil group’. Since then, deeper observations have found thissystem to consist of galaxies other than the bright centralgalaxy (BCG; Jones et al. 2000) and as a result the magni-tude gap criterion of fossils has been established. The mo-tivation for this criterion is that over time, an increasinglygrowing difference between the two brightest galaxies willform as a result of the merging of the most massive galax-ies into a single bright central elliptical if no infall occurs.This formation scenario is well suited for group mass fos-sils where the velocity dispersion is low and the dynamicalfriction time-scale is short.A number of objects meeting the fossil criteriahave also been observed in the cluster mass regimeas well (Cypriano et al. 2006; Khosroshahi et al. 2006;Voevodkin et al. 2010; Aguerri et al. 2011; Harrison et al.2012). It is possible fossil clusters may form as the resultof two systems merging, where one group has had its brightgalaxies merge due to dynamical friction, and the other hascomparatively fainter galaxies (Harrison et al. 2012). Shouldmerging occur between systems with similarly bright galax-ies, any previously existing magnitude gaps may becomefilled in. Therefore, meeting the fossil criteria may only bea transitory phase in the evolution of a group or cluster(von Benda-Beckmann et al. 2008; Dariush et al. 2010).Numerical and hydrodynamic simulations indicate thelarge magnitude gaps characterizing fossil groups and clus-ters are associated with an early formation time: fossilsystems have been found to assemble more of their totaldynamical mass than non-fossil systems at every redshift(Dariush et al. 2007), where half the dynamical mass is as-sembled by z & (D’Onghia et al. 2005). Evidence thatfossils have formed and evolved in a different manner thannormal groups and clusters should then manifest in differ-ences in their respective properties.The bright central galaxy which dominates the opticaloutput of fossil systems has a number of unique character-istics, although whether this demonstrates a clearly distinctformation scenario from non-fossil BCGs is still uncertain.The BCGs of fossils are more massive in both the stellarcomponent and in total than the central ellipticals in non-fossil systems of the same halo mass (Harrison et al. 2012).Méndez-Abreu et al. (2012) find fossil BCGs are consistentwith the Fundamental Plane of non-fossil BCGs, but showlower velocity dispersions and higher effective radii whencompared to non-fossil intermediate-mass elliptical BCGs ofthe same K s -band luminosity. These results suggest the fos-sil BCG has experienced a merger history of early gas-richdissipational mergers, followed by gas-poor dissipationlessmergers later. On the global scale, the scaling relations of fossil sys-tems remain a point of contention due to limited data andinhomogeneities between studies. Khosroshahi et al. (2007,hereafter KPJ07) performed a comprehensive analysis of asample of group mass fossil systems and found their samplefell on the same L X – T X relation as non-fossils. However, thefossil groups were found to have offset L X and T X for a givenoptical luminosity L opt or velocity dispersion σ v when com-pared to normal groups, which was interpreted as an excessin the X-ray properties of fossil systems for their mass. In acomparable study, Proctor et al. (2011) found similar devi-ations between fossils and non-fossils. This offset, however,was interpreted as fossils being underluminous in the opti-cal which is supported by their large mass-to-light ratios.These features would not result from galaxy-galaxy mergingin systems with normal luminosity functions, and thus thisanalysis calls into question the formation scenario commonlyattributed to generating the characteristic large magnitudegap of fossil systems. Later studies, such as Harrison et al.(2012) and Girardi et al. (2014, hereafter G14), find no dif-ference in the L X – L opt relation of fossil systems and non-fossils. Even so, most recently Khosroshahi et al. (2014)present a sample of groups, one of which qualifies as a fossil,that lies above the L X – L opt relation of non-fossil systems,reopening the debate on fossil system scaling relations.In this paper we have undertaken an X-ray study of10 candidate fossil systems, never previously studied withdetailed pointed observations in the X-ray regime. Using Suzaku data, we present the first measurements of intra-cluster medium (ICM) temperatures, bolometric X-ray lu-minosities, and estimates of the M masses of our sys-tems. This work comprises the sixth instalment of the FOs-sil Group Origins (FOGO) series. The FOGO project isa multiwavelength study of the Santos et al. (2007) can-didate fossil system catalogue. In FOGO I (Aguerri et al.2011), the FOGO project is described in detail and thespecific goals of the collaboration are outlined. FOGO II(Méndez-Abreu et al. 2012) presents a study of the BCGscaling relations of fossil systems and the implications forthe BCG merger history. Global optical luminosities ofour FOGO sample are measured in FOGO III (G14) andused to construct the global L X – L opt relation which re-veals no difference between the fossil and non-fossil fits.Deep r -band observations and an extensive spectroscopicdatabase were used to redetermine the magnitude gaps ofthe FOGO sample and reclassify our fossil candidate cata-logue in FOGO IV (Zarattini et al. 2014, hereafter Z14). InFOGO V (Zarattini et al. 2015), the correlation of the sizeof the magnitude gap and the shape of the luminosity func-tion is investigated. In this work (FOGO VI) we advance thecharacterization of the X-ray properties of fossil systems andconstrain the global scaling relations of these objects.The details and observations of our Suzaku sample aredescribed in Sections 2 and 3. A discussion on how non-ICM sources may contribute to the observed emission of oursystems follows in Section 4. Tests to determine the contri-bution of these non-ICM sources are presented in Sections 5and 6. Measurements of the global ICM properties of thethermally dominated subset of our sample are recorded inSection 7. Global scaling relations and their implications arepresented in Section 8. For our analysis, we assume a Λ CDMcosmology with a Hubble parameter H =70 km s − Mpc − , c (cid:13) , 1–17 ossil group origins – VI a dark energy density parameter of Ω Λ =0.7, and a matterdensity parameter Ω M =0.3. Our sample of 10 observed galaxy groups and clusters was se-lected from the Santos et al. (2007, hereafter S07) catalogueof candidate fossil systems. The S07 catalogue was assem-bled by first identifying luminous r <
19 mag red galaxies inthe luminous red galaxy (LRG) catalogue (Eisenstein et al.2001), and selecting only those galaxies associated withextended X-ray emission in the
ROSAT
All-Sky Survey(RASS). Sloan Digital Sky Survey (SDSS) Data Release 5was then used to spatially identify companion galaxies tothese bright galaxies. Group or cluster membership was as-signed to galaxies identified within a radius of 0.5 h − Mpcfrom one of the bright LRGs and with a redshift consistentwith that of the LRG. While spectroscopic redshifts wereused when available, galaxy membership was primarily de-termined using photometric redshifts. Groups and clusterswith more than a 2 r -band magnitude difference betweenthe brightest and second brightest member galaxies withinthe fixed 0.5 h − Mpc system radius were then selected, andthose with an early-type BCG were identified as fossils.Z14 observed the S07 fossil candidate list with theNordic Optical Telescope, the Isaac Newton Telescope, andthe Telescopio Nazionale Galileo to obtain deeper r -band im-ages and spectroscopic redshifts for candidate group mem-bers allowing for improved system membership. Addition-ally, the search radius for galaxy system members was ex-tended to the virial radius of the system as calculated fromthe RASS X-ray luminosity. The Z14 study confirms 15 tar-gets out of 34 S07 candidates are fossil galaxy systems. Ac-cording to this characterization, our sample contains fiveconfirmed fossil systems and five non-confirmed or rejectedfossil systems (see Table 1). The 10 systems in our sample were observed with the
Suzaku
X-ray telescope between 2012 May and October (Ta-ble 1). Our analysis uses the data from
Suzaku ’s three X-ray Imaging Spectrometers (XIS) sensitive to the 0.5–10keV band. Our single-pointing observations were taken witha normal clocking mode, and an editing mode of 3 × × HEASOFT version 6.15 software library with the calibra-tion database
CALDB
XIS update version 20140520. Spec-tra were extracted using
XSELECT version 2.4c and fit us-ing
XSPEC version 12.8.1g. The event files were reprocessedusing aepipeline with the
CALDB
XIS update 20140203using the default settings with an additional criterion ofCOR >
6. In our spectral analysis, emission from the Fecalibration sources, located in the corners of each XIS de-tector, was removed. Additionally, the XIS0 damaged pixelcolumns caused by micro-meteorites were masked.A Redistribution Matrix File (RMF) was created for all spectral extraction regions with xisrmfgen . For each RMF,two Ancillary Response Files (ARFs) were created with xissimarfgen , one to be convolved with the backgroundspectral model, and the other to be convolved with thesource model following the method of Ishisaki et al. (2007).Background ARFs were created out to a radius of 20 arcminusing a uniform emission source mode. For the source ARFs,an image of the stacked XIS field-of-view (FOV) was usedto model the emission.
High fidelity measurements of the ICM temperature and lu-minosity require careful consideration of non-ICM sourcesof emission during our analysis.
The standard
Suzaku
XIS background consists of a non-X-ray particle background (NXB; Tawa et al. 2008), the cos-mic X-ray background (CXB; Fabian & Barcons 1992), andforeground Galactic emission from the Local Hot Bubble(LHB) and the Milky Way Halo (MWH; Kuntz & Snowden2000).The contribution of the NXB for each object was as-sessed using the night earth database within 150 days ofthe observation using the FTOOL xisnxbgen (Tawa et al.2008). Our XIS1 observations were taken in a charge injec-tion mode of CI = 6 keV which increases the NXB. Accord-ingly, the nxbsci6 calibration file was used as input forXIS1 to counteract this.The contribution of the galactic foreground to aXIS spectrum is well described by two thermal plasmamodels: apec
LHB +( wabs × apec MWH ) where z LHB = z MWH = 0 , Z LHB = Z MWH = 1 Z ⊙ , and kT LHB =0.1 keV (Kuntz & Snowden 2000). The CXB was mod-elled by an absorbed power-law: wabs × powerlaw CXB with
Γ = 1 . (Kushino et al. 2002). During spec-tral analyses, the summed background and foregroundmodel: apec LHB + wabs ( apec MWH + powerlaw CXB ) wasconvolved with the uniform emission ARF. The interaction of ions in the solar wind with neutral atomsin the heliosphere and in Earth’s atmosphere can produce
E < keV photons in the X-ray regime (Cravens 2000;Fujimoto et al. 2007). To check for contamination from so-lar wind charge exchange (SWCX), proton flux light curveswith a sampling frequency of 90 s were obtained from theNASA WIND-SWE database over the time span of each ob-servation. The intensity of proton flux has been found to berelated to the strength of geocoronal SWCX contaminatingphotons, where flux levels above × protons cm − s − commonly indicate potentially significant contamination toX-ray spectra from charge exchange (Yoshino et al. 2009).Following Fujimoto et al. (2007), 2700 s were added to thetime points in the WIND-SWE light curve to account forthe travel time between the WIND satellite, located at theL1 point, and Earth, where the geocoronal SWCX emissionis produced. c (cid:13) , 1–17 A. Kundert et al.
Table 1.
Summary of observations.Object Sequence number RA Dec. Start date Exposure [ks] Type † FGS03 807052010 07:52:44.2 +45:56:57.4 2012 Oct 28 18:39:14 14.3 FFGS04 807053010 08:07:30.8 +34:00:41.6 2012 May 06 16:24:20 10.1 NCFGS09 807050010 10:43:02.6 +00:54:18.3 2012 May 30 05:18:38 9.9 NCFGS14 807055010 11:46:47.6 +09:52:28.2 2012 May 29 17:06:08 12.4 FFGS15 807057010 11:48:03.8 +56:54:25.6 2012 May 26 17:58:41 13.6 NFFGS24 807058010 15:33:44.1 +03:36:57.5 2012 Jul 28 08:10:10 13.2 NFFGS25 807049010 15:39:50.8 +30:43:04.0 2012 Jul 28 18:06:02 10.6 NFFGS26 807054010 15:48:55.9 +08:50:44.4 2012 Jul 29 02:05:54 8.6 FFGS27 807056010 16:14:31.1 +26:43:50.4 2012 Aug 05 07:14:36 10.6 FFGS30 807051010 17:18:11.9 +56:39:56.1 2012 May 02 11:43:31 14.0 F † The fossil status column contains the Z14 updated fossil characterizations of the S07 catalogue. In the fossilstatus column, ‘F’ is a confirmed fossil, ‘NF’ is a rejected fossil, and ‘NC’ is not confirmed as either a fossilor non-fossil according to Z14 and remains a fossil candidate.
Much of the FGS24 observation occurs during an el-evated period of proton flux; however, the light curve ofFGS24 displays no significant duration flares. Furthermore,as a check, we have performed our spectral analysis onthe time windows where the proton flux was less than × cm − s − and found the results were consistent withthe spectral analysis of the full baseline. We therefore con-sider the effects of SWCX to be small and have recorded theresults of the analysis of the full observation in the main textand include the FGS24 light curve and shortened exposuretime analysis in Appendix A. Our
Suzaku observations are the first pointed X-ray ob-servations of the objects in our sample. Consequently wemust assess point source contamination primarily relying onthe
Suzaku data alone. Visual inspection of the XIS images(Fig. 1) reveal two obvious point sources in the FGS15 FOVwhich we are able to exclude in our analysis using circular re-gions of radius 2.5 arcmin. Additionally, FGS03 and FGS09show diffraction spikes from a strong point-like sources nearthe peak of the X-ray emission. However, the large 2 arcminhalf-power diameter (HPD) of the
Suzaku
X-ray Telescope(XRT; Serlemitsos et al. 2007) inhibits the exclusion of thesesources and the robust identification of other point sources.Optical and radio studies of the objects in our samplehave found a number of active galactic nuclei (AGN) in spa-tial proximity to our galaxy systems. Especially concerningare the radio-loud AGN, located near the projected locationof the BCGs, found in 7 out of the 10 objects in our sample(Hess et al. 2012). To determine if these radio-loud AGN,and other optical and radio AGN in the FOV, are signifi-cant contributors to the source emission in the X-ray regime,we perform image (Section 5) and spectral (Section 6) anal-yses. In the 0.5–10 keV range of the XIS, the strength ofAGN emission increases towards the harder energies of thespectrum. As a result, the harder photons from an AGNmay falsely boost the measured temperature of the ICM ifonly a thermal model is used to fit the spectrum. AssessingAGN contribution is therefore a crucial step in determiningthe properties of the ICM.
Because most of our objects extend over the entire sin-gle
Suzaku pointing, a local
Suzaku background region isnot consistently available to assess the background contam-ination in our source regions. To aid in constraining theLHB, MWH, and CXB, we employ RASS background spec-tra sensitive to the 0.1–2.4 keV X-ray regime. RASS spec-tra were obtained through the High Energy AstrophysicsScience Archive Research Center (HEASARC) X-ray back-ground tool in an annulus of inner radius 0.5 degrees andouter radius 1 degree centred on each of our sources. Thesize of this annulus is sufficient to minimize contaminationfrom the source itself where the largest r radius found foran object in our sample only extends to ∼
20 per cent of theinner radius of the annular RASS background region.
The region of our initial spectral analysis for each object wasestablished to encircle where the emission from the sourcedominates the emission from the background, enabling theparameters describing the source spectrum to be determinedin a high signal-to-noise ratio (S/N) region. We determinethis source region using vignetting and exposure correctedimages of the source as well as simulated images of the back-ground estimated from RASS spectra.For each
Suzaku pointing, an exposure map was createdwith xisexpmapgen and a flat-field using xissim . The flat-field was simulated over the XIS 0.5–10 keV energy range ata monochromatic photon energy of 1 keV for a uniform skyout to 20 arcmin.An image of the NXB particle background for eachpointing was produced with xisnxbgen over the same energyrange. This image was estimated from night Earth observa-tions within 150 days of the
Suzaku observation date. TheNXB image was uniformly corrected by dividing the countrates by the exposure time.Emission from the CXB, LHB, and MWHwas estimated from RASS background spectra. http://heasarc.gsfc.nasa.gov/cgi-bin/Tools/xraybg/xraybg.plc (cid:13) , 1–17 ossil group origins – VI D e c li na t i on FGS03 D e c li na t i on FGS04 D e c li na t i on FGS09 D e c li na t i on FGS14 D e c li na t i on FGS15 D e c li na t i on FGS24 D e c li na t i on FGS25 D e c li na t i on FGS26 D e c li na t i on FGS27 D e c li na t i on FGS30
Figure 1.
The
Suzaku combined raw counts XIS0+XIS1+XIS3 images in the 0.5–10 keV band. The image is Gaussian smoothed with σ = 0.42 arcmin. White circlesdemarcate the initial spectral extraction region r ap , src defined to encircle the source-dominated region ( r ap , src values in Table 4). Fe calibration source events have beenremoved.
These spectra were fit with the background model: apec
LHB + wabs ( apec MWH + powerlaw CXB ) . Because theRASS background spectrum consists of only 7 data points,only the normalizations of the three background compo-nents were allowed to vary; the other parameters were fixed at the standard literature values as described in Section 4.1.The ROSAT
PSPC response matrix provided by thebackground tool was implemented for the fit. In calculatingthe background photon flux in the
Suzaku
XIS 0.5–10 keVenergy range, the
XSPEC dummyrsp command was used to c (cid:13) , 1–17 A. Kundert et al.
Table 2.
General information.FGS a Coordinates of Peak X-ray b z c n H d RA Dec. [10 cm − ] ∗ ∗ ∗ ∗ ∗ a [SMS2007] ID b Coordinates determined from the stacked XIS0+XIS1+XIS3raw count image in the 0.5–10 keV band c Spectroscopic redshift of the central bright galaxy in the fossilcluster (S07) d Weighted average galactic hydrogen column density in thedirection of the target (Kalberla et al. 2005) * Confirmed fossil system extrapolate beyond the
ROSAT
PSPC sensitivity range of0.1–2.4 keV.An image of the estimated CXB+LHB+MWH emissionwas produced with xissim out to a radius of 20 arcmin fromthe coordinates of the X-ray centre of the systems. The emis-sion was modeled with the best-fitting spectral model andphoton flux of the RASS background data. Because of thelow count rate of CXB+LHB+MWH photons over the ex-posure time for each object, the exposure time was increasedby a factor of 10, and corrected later, to improve the statis-tics of the surface brightness profile of the resulting imagefollowing the method of Kawaharada et al. (2010).An image of the source could then be created fromthe images constructed during this procedure. Because theNXB background is not affected by vignetting, the expo-sure corrected image of the NXB was subtracted from theexposure corrected image of the XIS detector. The result-ing image was then vignetting corrected with the flat-fieldand the vignetting and exposure corrected image of theCXB+LHB+MWH was subtracted to obtain the estimatedvignetting corrected image of source emission.Surface brightness profiles were created using ds9 for thevignetting corrected source, NXB, and CXB+LHB+MWHimages as shown for example in Fig. 2. The coordinates ofpeak X-ray emission (Table 2) were used as the centre ofthe surface brightness profile. The profile was constructedfrom 20 uniformly spaced circular annuli out to the radiusof the largest circle that could be inscribed within the XISFOV from the centre coordinates. The source and combinedbackground profiles were then averaged for the three XISdetectors and the radius at which the source and backgroundemission are equal was identified. We find that within thisradius the source contributes on average ∼
80 per cent ofthe total counts, with no less than a ∼
70 per cent sourcecontribution for all objects in our sample. It is this radius,the source radius r ap , src , which we have used to define ourregion of initial source spectral analysis. -10 -9 -8 -7 -6 s u r f a c e b r i g h t n e ss [ c t s s − a r c s e c − ] XIS0
FGS30 sourceCXB+GXE+NXBCXB+GXENXB 0 50 100 150 200 250 300 350 400 450radius (arcsec)10 -9 -8 -7 -6 s u r f a c e b r i g h t n e ss [ c t s s − a r c s e c − ] XIS10 50 100 150 200 250 300 350 400 450radius (arcsec)10 -10 -9 -8 -7 -6 s u r f a c e b r i g h t n e ss [ c t s s − a r c s e c − ] XIS3 0 50 100 150 200 250 300 350 400 450radius (arcsec)10 -9 -8 -7 -6 s u r f a c e b r i g h t n e ss [ c t s s − a r c s e c − ] XIS combined r ap,opt =5.4'
Figure 2.
An example of the estimated source and backgroundsurface brightness profiles for FGS30. The bottom right-handpanel shows the average source and background profile for thethree XIS detectors.
Radial surface brightness profiles were constructed for eachobject using stacked 0.5–10 keV XIS0+XIS1+XIS3 observedimages. For the purpose of this profile analysis, we applyan additional satellite attitude correction to the event filesused to create the images.
Suzaku
XIS images can containup to a 1 arcmin position error as a result of a recurrent off-set between the XRT optical axis and the satellite attitude(Uchiyama et al. 2008). With the application of a correctedattitude file, the XIS images can thus be sharpened. Thiscorrection was performed by generating corrected attitudefiles with aeattcor , and then applying these corrected at-titude files to our cleaned event files using xiscoord . Thenew corrected event files are used to produce the imagesused in our brightness profile analysis, the brightness pro-files of which are shown in Fig. 3. The number of annulifor each profile was determined such that each annulus hadat minimum 225 counts, which, assuming Poissonian noise,requires the number of counts to be 15 times the error.The brightness profile of a spherically symmetric andisothermal ICM in hydrostatic equilibrium will follow a β -model (Cavaliere & Fusco-Femiano 1976, 1978). These areappropriate assumptions for virialized and relaxed groupsand clusters. Disparity between the data and the sin-gle β -model can therefore result from processes such asmerger asymmetries, multiple thermal components, andnon-thermal emission, for example, as produced by an AGN.Our initial fit of the profiles consists of a β -model plus abackground constant: S ( r ) = S (1 + ( r/r c ) ) − β +1 / + k, (1)where S is the central surface brightness, r c is the coreradius, and k is the background surface brightness. In thismodel, the β -model component was convolved with a ra-dial model of the Suzaku
XRT PSF (see Appendix B). Fits c (cid:13) , 1–17 ossil group origins – VI Table 3.
Best-fitting parameters of the surface brightness pro-files.
FGS β -model + background constant S † r c β k † χ /d.o.f. ( χ )[10 − ] [kpc] [10 − ]03 ∗ +32 . − . +2 − + ∞− . +1 . − . + ∞− . +7 − +0 . − . +0 . − . +4 . − . +5 − + ∞− . +0 . − . ∗ +2 . − . +11 − +0 . − . +0 . − . + ∞− . +3 − +0 . − . +0 . − . +0 . − . +19 − +0 . − . +0 . − . +2 . − . +3 − +0 . − . +1 . −∞ ∗ +0 . − . +7 − +0 . − . +5 . −∞ ∗ +0 . − . +22 − +0 . − . +0 . − . ∗ +16 . − . +3 − +0 . − . +2 . − . † Units of counts s − Mpc − * Confirmed fossil system were performed with the Sherpa Python module (Doe et al.2007).The returned best-fitting parameters are recorded inTable 3 and the convolved best-fitting model is shown inFig. 3. We note that FGS03, FGS09, FGS15 have χ r > indicating the β -model poorly describes the observed emis-sion. For these objects, we test adding to the original modela point-like component consisting of a δ function convolvedwith the PSF model. This additional point-source compo-nent does not offer an improvement in χ r compared to theoriginal β -model fits. Nevertheless, the emission from thesethree objects seems to indicate that either the ICM is notrelaxed, or there is some significant source of non-ICM emis-sion.Because the annuli used are smaller than the Suzaku
XRT PSF and, additionally, discrepancy from a β -modelcould be attributed to multiple phenomena, we consider theresults as merely suggestive and to be used and interpretedin conjunction with our spectral analysis. Our spectral analysis consists of measuring spectral proper-ties within a region of high S/N (Section 6.1) and using theseresults to classify these objects as thermally dominated orAGN contaminated (Section 6.2). The results of this sectionwill then be used to measure or estimate the global proper-ties of the ICM-dominated systems within r (Section 7). In order to disentangle ICM emission from potential con-taminating point source emission, we perform our analysison the source aperture region where the source emission ismore than half of the total emission from the object. By de-termining this source aperture radius, r ap , src as described inSection 5.1, we make no assumptions on the type of sourceemission. Extracting a spectrum from this region thereforeimproves the spectral analysis of any type of source overthe background whether the source is dominated by ther-mal emission from the ICM or non-thermal emission froman AGN.The results of our surface brightness profile analysis in-dicate some objects in our sample may have a strong non-thermal point-like component to the total emission. As a result, we compare the fit of three source models to ourspectra:(i) an absorbed thermal plasma model, wabs × apec , tomodel the ICM;(ii) an absorbed power-law, wabs × powerlaw , to modelan AGN;(iii) an absorbed combined thermal and power-law model, wabs ( apec + powerlaw ), to describe contributionfrom both the ICM and an AGN;where the wabs absorption component accounts for galacticabsorption in all three models.The background and foreground sources consist of theNXB, LHB, MWH, and CXB. The NXB spectrum was usedas the background file for the extracted r ap region to be sub-tracted directly during the spectral fit. The CXB, LHB, andMWH were accounted for through modelling as described inSection 4.1.The XIS spectra were grouped with grppha such thateach bin had a minimum of 25 counts. The binned Suzaku
XIS0, XIS1, and XIS3 spectra were fit simultaneously withthe RASS background spectrum. The
Suzaku spectra werefit with the source and background model while the RASSspectra were fit only with the background model. The RASSbest-fitting parameters were tied to that of the
Suzaku spec-tra with a scaling factor to account for the difference in theangular size of the spectral extraction regions. Bad channelswere ignored for all spectra. The
Suzaku
XIS0 and XIS3spectra were fit over 0.7–10 keV (Section 6.1.1), the XIS1spectra over 0.7–7 keV, and the RASS spectra over the range0.1–2.4 keV.In all three models, the neutral hydrogen column den-sity was assigned the weighted average galactic value in thedirection of the source (Kalberla et al. 2005). The redshiftsof our systems were taken to be the spectroscopic redshiftsof the bright central galaxies as determined by S07. Duringthe fit, the column density and redshift were always fixed.The metal abundance Z component of the apec model wascalculated using the abundance tables of Anders & Grevesse(1989). The photon index of the powerlaw model was con-strained to be within Γ = 1 . − . (Ishibashi & Courvoisier2010). Initially, all other parameters were left free to be fit.However, if during the fit convergence on an apec or pow-erlaw parameter within the physically reasonable limits didnot occur or the parameter was returned with infinite errorbars, the fit was performed again with that parameter fixed.In all further tables, quantities presented without error barshave been fixed to a reasonable value.The resulting best-fitting parameters are listed in Ta-ble 4 and the best-fitting models to the spectra are shownin Fig. 4. The background parameters resulting from eachof the model fits were consistent with each other within 1 σ errors. While the XIS is sensitive to photons with energy as low as0.5 keV, we have excluded the
E < . keV energy channelsfrom our spectral analysis. In the majority of our observa-tions, an apparent excess in counts was found in the 0.5–0.7keV range when compared to the fit of the apec or power-law models in the E > . keV range. c (cid:13) , 1–17 A. Kundert et al. radius [arcsec]10 -4 -3 -2 -1 s u r f a c e b r i g h t n e ss [ c t s s − M p c − ] FGS03 d a t a / m o d e l radius [arcsec]10 -4 -3 -2 -1 s u r f a c e b r i g h t n e ss [ c t s s − M p c − ] FGS04 d a t a / m o d e l radius [arcsec]10 -4 -3 -2 -1 s u r f a c e b r i g h t n e ss [ c t s s − M p c − ] FGS09 d a t a / m o d e l radius [arcsec]10 -4 -3 -2 -1 s u r f a c e b r i g h t n e ss [ c t s s − M p c − ] FGS14 d a t a / m o d e l radius [arcsec]10 -4 -3 -2 -1 s u r f a c e b r i g h t n e ss [ c t s s − M p c − ] FGS15 d a t a / m o d e l radius [arcsec]10 -4 -3 -2 -1 s u r f a c e b r i g h t n e ss [ c t s s − M p c − ] FGS24 d a t a / m o d e l radius [arcsec]10 -4 -3 -2 -1 s u r f a c e b r i g h t n e ss [ c t s s − M p c − ] FGS25 d a t a / m o d e l radius [arcsec]10 -4 -3 -2 -1 s u r f a c e b r i g h t n e ss [ c t s s − M p c − ] FGS26 d a t a / m o d e l radius [arcsec]10 -4 -3 -2 -1 s u r f a c e b r i g h t n e ss [ c t s s − M p c − ] FGS27 d a t a / m o d e l radius [arcsec]10 -4 -3 -2 -1 s u r f a c e b r i g h t n e ss [ c t s s − M p c − ] FGS30 d a t a / m o d e l Figure 3.
Surface brightness profiles of the stacked XIS image in the 0.5–10keV band. The best-fitting convolved β -model is plotted in solid red; dashed linesrepresent the components to the model. Residuals for the β -model are plotted assquares. Potential origins of this soft excess include a secondthermal component in the ICM, an AGN, calibration issues,SWCX, or statistical fluctuations. Adding a second thermalmodel to the ICM model did not improve the fit. If an AGNwere the origin of the excess, removing the softest energies should not greatly deter detecting the presence of its emis-sion in the spectra because an AGN will contribute moststrongly to the harder energies of the spectrum. Calibrationissues with proportional removal of flickering pixels from ob-servations of the source and the NXB may also contribute c (cid:13) , 1–17 ossil group origins – VI to energy channels below 0.6 keV. Additionally, it is possi-ble there is some contribution from SWCX in the soft energyregime, although the solar wind proton flux light curves ofmost of our sample are of a low intensity indicating geocoro-nal SWCX emission is unlikely to be a significant contami-nant (see Section 4.2).Because the origin of this excess is uncertain and thuscannot be appropriately modelled in the spectra, and fur-thermore the excess only affects the first few low energychannels in the spectrum, we exclude this softest energyregime from our fits. This has little effect on the returnedbest-fitting parameters and in general the reduced χ of thefits improves with the exclusion of the soft excess energychannels. In comparing the fits of the three models, the apec + powerlaw model does not appear to signifi-cantly improve the characterization of the spectra overthe individual apec and powerlaw fits. Indeed in thecombined fit, the apec and powerlaw components arenever simultaneously constrained. As a result, while some apec + powerlaw fits return χ r with values slightly lessthan that for the less complex fits of apec or powerlaw only, we decide to choose the simpler model that has allparameters constrained.By the χ r values, the powerlaw model provides a bet-ter fit over the thermal apec model for FGS03, FGS09,FGS15, and FGS24. We consider these four objects to bedominated by a non-ICM source and with our current ob-servations, we cannot disentangle the ICM and non-ICMemission. Further discussion on these objects is provided inAppendix C.For the remainder of our analysis, we focus on thoseobjects in our sample with spectra that are best fit by the apec model and are thus galaxy systems dominated by ICMemission. In order to compare the ICM temperatures and luminositiesof our fossil systems with those of other groups and clusters,we calculate these properties within the fiducial radius of r , the radius at which the average enclosed density is500 times the critical density of the Universe. We calculate r , and the spectral properties within this radius, using aniterative procedure.Using the temperature calculated within some aper-ture, T ap , we calculate r using the r – T X relation ofArnaud et al. (2005): r = 1 . h − E ( z ) − (cid:18) kT (cid:19) . Mpc , (2)where h = H / (70 km s − Mpc − ) and E ( z ) = H ( z ) /H = p Ω M (1 + z ) + Ω k (1 + z ) + Ω Λ (Hogg 1999).This value of r is used as our next radius of extraction todetermine a new T ap , and we continue this process until con-vergence is reached between r and the temperature, and T a b l e . B e s t - fi tt i n g s p ec t r a l p a r a m e t e r s i n t h e r a p , s r c r e g i o n . F G S r a p , s r c a pe c p o w e r l a w a pe c + p o w e r l a w k T a p ec Z a p ec n o r m a a p ec χ / d . o . f ( χ r ) Γ P L n o r m b P L χ / d . o . f ( χ r ) k T a p ec Z a p ec n o r m a a p ec Γ P L n o r m b P L χ / d . o . f ( χ r ) [’ ( M p c ) ] [ k e V ][ Z ⊙ ][ − ] [ − ] [ k e V ][ Z ⊙ ][ − ] [ − ] ∗ . ( . ) . + . − . . + . − . . + . − . ( . ) . + . − . . + . − . ( . ) . . + . − . . + . − . . + . − . ( . ) . ( . ) . + . − . . + . − . . + . − . ( . ) . + . − . . + . − . ( . ) . + . − . . + . − . . + . − . . . + . − . ( . ) . ( . ) . + . − . . + . − . . + . − . ( . ) . + . − . . + . − . ( . ) . . + . − . . + . − . . + . − . ( . ) ∗ . ( . ) . + . − . . + . − . . + . − . ( . ) . + . − . . + . − . ( . ) . + . − . . + . − . . + . − . . . + . − . ( . ) . ( . ) . + . − . . . + . − . ( . ) . + . − . . + . − . ( . ) . . + . − . . . + . − . ( . ) . ( . ) . + . − . . + . − . . + . − . ( . ) . + . − . . + . − . ( . ) . + . − . . . + . − . . . + . − . ( . ) . ( . ) . + . − . . + . − . . + . − . ( . ) . + . − . . + . − . ( . ) . + . − . . + . − . . + . − . . . + . − . ( . ) ∗ . ( . ) . + . − . . + . − . . + . − . ( . ) . + . − . . + . − . ( . ) . + . − . . + . − . . + . − . . . + . − . ( . ) ∗ . ( . ) . + . − . . + . − . . + . − . ( . ) . + . − . . + . − . ( . ) . + . − . . + . − . . + . − . . . + . − . ( . ) ∗ . ( . ) . + . − . . + . − . . + . − . ( . ) . + . − . . + . − . ( . ) . + . − . . + . − . . + . − . . . + . − . ( . ) N o t e : q u a n t i t i e s w i t h o u t e rr o r s h a v e b ee n fi x e d a tt h e li s t e d v a l u e a n o r m a p ec = −
14 4 π [ D A ( + z ) ] R n e n H d V c m − b n o r m P L h a s un i t s o f ph o t o n s k e V − c m − s − a r c m i n − a t k e V * C o n fi r m e d f o ss il s y s t e m c (cid:13) , 1–17 A. Kundert et al. −4 −3 no r m a li z ed c oun t s s − k e V − FGS 03 (powerlaw)1 2 50.511.52 r a t i o Energy (keV) 10 −3 no r m a li z ed c oun t s s − k e V − FGS 04 (apec)1 2 511.5 r a t i o Energy (keV) 10 −3 no r m a li z ed c oun t s s − k e V − FGS 09 (powerlaw)1 2 50.511.52 r a t i o Energy (keV)10 −3 no r m a li z ed c oun t s s − k e V − FGS 14 (apec)1 2 511.52 r a t i o Energy (keV) 10 −3 no r m a li z ed c oun t s s − k e V − FGS 15 (powerlaw)1 2 511.5 r a t i o Energy (keV) 10 −3 no r m a li z ed c oun t s s − k e V − FGS 24 (powerlaw)1 2 512 r a t i o Energy (keV)10 −3 no r m a li z ed c oun t s s − k e V − FGS 25 (apec)1 2 51234 r a t i o Energy (keV) 10 −3 no r m a li z ed c oun t s s − k e V − FGS 26 (apec)1 2 50.511.522.5 r a t i o Energy (keV) 10 −4 −3 no r m a li z ed c oun t s s − k e V − FGS 27 (apec)1 2 511.52 r a t i o Energy (keV)10 −3 no r m a li z ed c oun t s s − k e V − FGS 30 (apec)1 2 50.511.5 r a t i o Energy (keV)
Figure 4.
The XIS0 (black), XIS1 (red), and XIS3 (green) spectra for the source regions r ap , src determined in Section 5.1. The best-fitting model to the observed spectra, asdetermined by the χ values in Table 4, is plotted in a solid line. The RASS spectrathat were simultaneously fit with the Suzaku background model are not shown. thus T has been determined. This analysis is performedon the subset of our sample that is thermally dominated(Section 6.2). The iterative process is begun with the T ap determined from the apec only fit as recorded in Table 4.For two of our objects, FGS25 and FGS26, the finalestimation of r extends beyond the largest aperture ra-dius that can be inscribed within the XIS FOV. However,our estimated r is very similar to the largest aperturesize that was used to extract spectral parameters, wherethe ratio between the maximum r ap and r is 0.98 and0.84 for FGS25 and FGS26, respectively. As a result the T ap values for these two objects should reasonably describe thetrue global temperature within r . When considering theluminosity, L X , is estimated from L X , ap using a surface brightness profile model that well describes the ICM emis-sion. By integrating this surface brightness model over area,the conversion factor between L X , and L X , ap is calculatedusing the relation L X , L X , ap = R r S ( r ) r dr R r ap S ( r ) r dr (3)where S is an azimuthally averaged radial surface brightnessprofile. For our surface brightness model, we use the β -modelwhere S , r c , and β have the values recorded in Table 3.With the global temperature values listed in Table 5,we estimate the masses within r for our systems using c (cid:13) , 1–17 ossil group origins – VI the M – T X relation of Arnaud et al. (2005): M = 3 . × h − E ( z ) − (cid:18) kT (cid:19) . M ⊙ . (4)We find our thermally dominated objects have masses con-sistent with clusters ( M & M ⊙ ). We combine our newly measured global L X , bol , and T X with previously measured fossil systems properties, to con-strain the scaling relations of these objects with the goal ofassessing if fossil systems display different scaling relationsthan those for normal groups and clusters. Our analysis offossil system scaling relations is distinguished from previousstudies through several updates including the fitting of thelargest assembled fossil system data set, using recent X-rayand optical data for our control sample of normal groups andclusters, and a substantial effort of homogenizing both thefossil and non-fossil data sets. We furthermore record thebest-fitting L X – L r relation, and for the first time record theslopes and y -intercepts of the L X – T X , L X – σ v , T X – σ v scalingrelation fits for fossil systems. We have assembled data from a number of studies to inves-tigate how the global X-ray and optical properties of fossilsystems compare to non-fossil groups and clusters. To en-sure a reliable comparison, we have made an effort to usequantities determined within the same fiducial radius anddefined the same way. For our analysis we use bolometric X-ray luminosities L X , bol , temperatures T X , and optical SDSS r -band luminosities L r all calculated within r , and globalvelocity dispersions σ v . While we have removed known fos-sils from our sample of non-fossil groups and clusters, wedo not have information on the magnitude gap between thefirst and second brightest galaxies in all of the systems mak-ing up our control sample. However, the large magnitudegap characterizing fossil systems should be found in only afraction of L X , bol > × erg s − systems (Jones et al.2003; Milosavljević et al. 2006). Thus, we expect our controlsample is contaminated by at most a few unidentified fossilsystems.To assemble our group sample, we use the σ v of the‘G-sample’ from Osmond & Ponman (2004). Group T X val-ues are pulled from Rasmussen & Ponman (2007), Sun et al.(2009), Hudson et al. (2010), Eckmiller et al. (2011), andLovisari et al. (2015). Lovisari et al. (2015) L X , . − . are transformed to L X , bol using the conversion tables ofBöhringer et al. (2004).For the cluster sample, we use the G14 r -band op-tical luminosities calculated within r . The correspond-ing velocity dispersions are taken from Girardi et al. (1998,2002), Girardi & Mezzetti (2001), Popesso et al. (2007), andZhang et al. (2011). Bolometric X-ray luminosities within r and temperatures were sourced from Pratt et al. (2009)and Maughan et al. (2012), and supplemented with addi-tional L X , bol from Zhang et al. (2011) and T X from Wu et al.(1999) and Hudson et al. (2010).Taking our sample of fossil systems observed with Suzaku , we match the global X-ray properties of the sys-tems in Table 5 with the corresponding L r from G14 and σ v from Z14. For the remainder of the Z14 confirmed fossil cat-alogue, we supplement the L X , bol from G14. For improvedconsistency with the L X of our cluster sample, we approx-imate X-ray luminosities that more closely resemble thosecomputed using the growth curve analysis (GCA) method(Böhringer et al. 2000) from the G14 luminosities derivedfrom RASS counts (see section 3.3 of G14 for details). Thesecorrected luminosities also show good agreement with the Suzaku measured L X for the sample of objects shared be-tween both the G14 study and ours.We add to the fossil sample with the X-ray luminosi-ties and temperatures from KPJ07 and Miller et al. (2012),matched with the L r and σ v data from Proctor et al. (2011).The KPJ07 L X , bol , are rescaled to r using their best-fitting β -model parameters and our luminosity conversionEq. 3. To ensure consistency with our Suzaku sample, the r of KPJ07 is recalculated from their temperatures usingour Eq. 2 and we use this value to estimate L X , bol , . Torescale the L r, of Proctor et al. (2011) to r , we assumethe light follows the mass, which is a good approximationon the global scale of groups and clusters (Bahcall & Kulier2014). For a NFW density profile with concentration pa-rameter c = 5 , M /M = 0 . (Navarro et al. 1997).The assumption of c = 5 was chosen for agreement withthe concentrations of normal clusters of similar tempera-ture and mass (Pointecouteau et al. 2005; Pratt & Arnaud2005; Vikhlinin et al. 2006; Buote et al. 2007; Ettori et al.2010) because the typical concentration parameter for fossilsystems is poorly characterized. Thus, we can rescale us-ing L opt , /L opt , ∝ . . Here, the correction relation isapplied only to the non-BCG light.We also implement the fossil catalogue of Harrison et al.(2012). We take their L X , bol , and rescale by assum-ing a β -model with r c estimated using the r c – L X relationof Böhringer et al. (2000) and β =0.67, then correcting to L X , bol , using Eq. 3. The optical luminosities provided arecalculated for r = 0 . r ∼ r . By the reasoning de-scribed previously, this luminosity is corrected to L r, us-ing the relation M /M ∝ . . Because the magnitudesof the BCG were not recorded, we rescale all of the opticallight for these objects. The Harrison et al. (2012) σ v are alsoused, and we assign a 0.1 dex error to these values duringour fit of the fossil scaling relations.With the above data sets, we have enough data to as-semble and quantitatively compare the L X – T X , L X – σ v , L X – L r , T X – σ v scaling relations for a sample of fossils and acontrol sample of normal groups and clusters. We do not in-vestigate the T X – L r relation due to the small subsample ofour control population with both T X and L r measurements.We fit the equation log( Y ) = a + b log( X ) (5)to the data using the BCES orthogonal method(Akritas & Bershady 1996) which accounts for mea-surement errors in the data as well as intrinsic scatter inthe fitted relation. We choose to compare the fit of the fossilsample to a combined sample of groups and clusters (G+C)in the same parameter range as the fossil sample. For thefossil system data set we exclude NGC 6482 from KPJ07 c (cid:13) , 1–17 A. Kundert et al.
Table 5.
Global properties of the ICM-dominated subsample.FGS r ap /r kT ap Z ap L X , bol , ap r L X , bol , r500 M [keV] [Z ⊙ ] [10 erg s − ] [10 erg s − ] [10 M ⊙ ]04 1 2.81 +0 . − . +0 . − . +0 . − . +0 . − . ± ∗ +0 . − . +0 . − . +0 . − . +0 . − . ± +0 . − . +0 . − . +0 . − . +0 . − . ± ∗ +0 . − . +0 . − . +0 . − . +0 . − . ± ∗ +0 . − . +0 . − . +0 . − . +0 . − . ± ∗ +0 . − . +0 . − . +0 . − . +0 . − . ± Note: L X , bol is the unabsorbed X-ray luminosity in the 0.1-100 keV energy range * Confirmed fossil system and XMMXCS J030659.8+000824.9 from Harrison et al.(2012) as they are clear outliers.We cross-checked the results obtained with the BCESmethod with the IDL Astronomy library tool
LIN-MIX_ERR (Kelly 2007), a Bayesian fitting method for lin-ear regression. The plotted scaling relations and BCES fitsare shown in Fig. 5 and the best-fitting parameters of therelations are recorded in Table 6. Uncertainties on the BCESbest-fitting parameters are estimated using 10,000 bootstrapresamplings. For the
LINMIX_ERR fits, the quoted val-ues are the mean and the standard deviation of the posteriordistributions for the regression parameters. We investigatechanging the pivot point of the fits, i.e. rescaling the X and Y values in Eq. 5 by a constant, but no difference is foundin the returned fits.We find the BCES fits to the fossil sample are consistentwithin error to the combined groups and clusters fit for eachscaling relation investigated in this work. The LINMIX fos-sil and non-fossil fits are for the most part consistent within1 σ ; the y -intercepts of L X , bol – T X and the y -intercepts andslopes of L X – σ v are consistent within 2 σ . These slight dis-crepancies in the LINMIX fits are most likely due to inho-mogeneities in the data or the small sample size of both thefossil and control populations.The global properties involved in these scaling rela-tions: L X , T X , L opt , σ v , are determined predominantly bythe shape and depth of the potential well, and are thus well-documented proxies for the total mass of the system. Addi-tional important effects that determine the X-ray propertiesof the ICM include the entropy structure (Donahue et al.2006) and non-gravitational heating and cooling processes,such as can be caused by AGN or mergers. These factors canproduce dispersions in scaling relations between X-ray andoptical mass proxies. Finding no difference in the scaling re-lations between fossil and non-fossil groups and clusters thusindicates fossil systems are of similar mass as non-fossils,and on the global scale, the combined effect of mass, ICMentropy, and non-gravitational processes that have occurredin fossil systems are similar to the combined effect of thosethat have occurred in normal groups and clusters. Our result that fossils share the same L X – T X relation as non-fossil groups and clusters is consistent with previous stud-ies (KPJ07; Proctor et al. 2011; Harrison et al. 2012, G14).However, the comparison of optical and X-ray properties of fossil and non-fossil systems is a contentious issue in theliterature.The L X – L r , L X – σ v , T X – σ v scaling relation fits of ouranalysis show the relations of fossil systems are consistentwithin error to normal groups and clusters. This is in goodagreement with the findings of Harrison et al. (2012) andG14. G14 recorded the first quantitative values of their fitto the L X – L r relation and found no difference between fossilsystems ( L X ∝ L . ± . r ) and a sample of non-fossil clusters( L X ∝ L . ± . r ). While qualitatively we both find no dif-ference in the L X – L r fossil and non-fossil scaling relations,there are some numerical differences in the returned best-fitting parameters of our study and G14.Our fossil fit of L X ∝ L . ± . r is consistent withinerror to G14, although this is in large part due to the con-siderable error on both of our slopes. However, our non-fossilfit ( L X ∝ L . ± . r ) is not within error of the fit determinedby G14. Differences in the slopes of our fits could be due tomultiple reasons: (1) we use bolometric L X in our fits, whileG14 uses L X , . − . ; (2) our L X are defined within r while the fitted G14 L X represent a total luminosity that hasnot been defined within a precise radius; (3) we use differentfitting methods; (4) we fit our control sample of non-fossilsover a different parameter space (i.e., one defined to matchour fossil sample).We check to see if we can return more consistent resultswith G14 by repeating our analysis of the L X – L r relationusing L X , . − . instead of L X , bol and expanding the fitof our control ‘G+C’ sample to the full parameter space. Wefind the returned fit to the fossil sample ( L X ∝ L . ± . r )and to the non-fossil sample ( L X ∝ L . ± . r ) are bothwithin error of the G14 fits. And again we emphasize thateven without the changes made here, although numericallyour fits differ from those of G14, the interpretation is thesame: fossil systems follow the same L X – L r scaling as non-fossil systems, supporting our conclusion that on the globalscale, fossil systems have optical and X-ray properties con-gruent with those of normal groups and clusters.Accumulation of multiple differences in data andmethodology explain the differences in conclusions betweenour study and those of earlier studies (KPJ07; Proctor et al.2011) that find discrepancies in the optical and X-ray scal-ing relations for fossil and non-fossils. We have comparedfossil and non-fossil optical luminosities measured from thesame photometric catalogue and band, avoiding the need tomake approximative luminosity estimates for comparisonsbetween samples. We have also used optical luminosities de-fined within the same fiducial radius, thus ensuring a more c (cid:13) , 1–17 ossil group origins – VI kT X [keV]10 L X , b o l [ e r g s / s ] groupsclustersclusters (K+)fossils (Z14/G14)fossils (M12/P11)fossils (KPJ07/P11)fossils (KPJ07)fossils (H12)fossils (K+)Fossil fitG+C fit σ v [km/s]10 L X , b o l [ e r g s / s ] L r [L r ⊙ ]10 L X , b o l [ e r g s / s ] σ v [km/s]10 k T X [ k e V ] Figure 5. L X , T X , L r , σ v scaling relations for fossil and non-fossil samples. We abbreviate this current work as K+, Zarattini et al.(2014) as Z14, Girardi et al. (2014) as G14, Miller et al. (2012) as M12, Proctor et al. (2011) as P11, Khosroshahi et al. (2007) as KPJ07,and Harrison et al. (2012) as H12. The plotted lines are the orthogonal BCES fits to the fossil sample (dashed line) and to the sampleof groups and clusters (solid line) in the same parameter range as the fossils. equal comparison between data pulled from multiple cata-logues. Additionally, our large sample size of fossils reducesthe effect of noise to ensure a more reliable comparison be-tween the fossil and non-fossil samples.We note, however, that our best-fitting parameters forboth the fossil and non-fossil samples have large errors.Thus, a study of fossil scaling relations could be greatly im-proved in the future by larger and more homogeneous data sets. Furthermore, our results probe the relations of clustersand high-mass groups, and consequently it is possible dif-ferences in the scaling relations exist in the low-mass end(Desjardins et al. 2014; Khosroshahi et al. 2014). c (cid:13) , 1–17 A. Kundert et al.
Table 6.
Best fits to the scaling relations.Relation ( Y - X ) Sample Fitting ProcedureBCES Orthogonal LINMIX_ERR a b a bL X , bol – T X Fossils 42.48 ± ± ± ± ± ± ± ± L X , bol – σ v Fossils 30.05 ± ± ± ± ± ± ± ± L X , bol – L r Fossils 15.98 ± ± ± ± ± ± ± ± T X – σ v Fossils -3.73 ± ± ± ± ± ± ± ± We have presented a detailed study of the X-ray propertiesof 10 candidate fossil galaxy systems using the first pointedX-ray observations of these objects. In particular,
Suzaku
XIS data have been used to measure their global X-ray tem-peratures and luminosities and to estimate the masses ofthese galaxy clusters. We determine 6 of our 10 objects aredominated in the X-ray by thermal bremsstrahlung emissionand thus we are able to measure the global temperaturesand luminosities of their ICM. This sample of six objectshas temperatures of . T X . , luminosities of . × L X , bol . × erg s − , and occupies thecluster regime in plotted scaling relations.Using our newly determined fossil cluster ICM X-rayproperties, we combine our fossil sample with fossils in theliterature to construct the largest fossil sample yet assem-bled. This sample is compared with a literature sample ofnormal groups and clusters, where significant effort has beenmade to homogenize the global L X , T X , L r , and σ v data forthe fossil and non-fossil samples.Plotting the L X – T X , L X – σ v , L X – L r , and T X – σ v rela-tions shows no difference between the properties of fossilsand normal groups and clusters. Furthermore, we providethe first fits to three of these relations which reveals the re-lations of fossils systems agree within error to the relationsof normal groups and clusters. Our work indicates that onthe global scale, fossil systems are no different than non-fossil systems. However, the distinguishing large magnitudegap in the bright end of the fossil system luminosity functionis still unexplained and thus further studies are necessary tocharacterize the properties of these objects and understandtheir nature. ACKNOWLEDGEMENTS
This research has made use of data obtained from the
Suzaku satellite, a collaborative mission between the space agen-cies of Japan (JAXA) and the USA (NASA). We thank theanonymous referee for valuable comments, K. Hamaguchiand K. Pottschmidt at the
Suzaku
Helpdesk for useful ad-vice on multiple aspects of our analysis, and D. Eckert forhelpful discussions and for suggesting the flickering pixelsissue as an explanation for the excess in the 0.5–0.7 keVrange.Support for this research was provided by NASA GrantNo. NNX13AE97G, and by the University of Wisconsin- Madison Office of the Vice Chancellor for Research andGraduate Education with funding from the WisconsinAlumni Research Foundation. FG acknowledges the finan-cial contribution from contract PRIN INAF 2012 (‘A uniquedataset to address the most compelling open questions aboutX-ray galaxy clusters’) and the contract ASI INAF NuS-TAR I/037/12/0. ED gratefully acknowledges the supportof the Alfred P. Sloan Foundation. MG acknowledges fund-ing from MIUR PRIN2010-2011 (J91J12000450001). JALAhas been partly funded from MINECO AYA2013-43188-P.EMC is partially supported by Padua University throughgrants 60A02-4807/12, 60A02-5857/13, 60A02-5833/14, andCPDA133894. JMA acknowledges support from the Euro-pean Research Council Starting Grant (SEDmorph; PI V.Wild).
REFERENCES
Aguerri, J. A. L., Girardi, M., Boschin, W., et al. 2011, A&A,527, A143Akritas, M. G., & Bershady, M. A. 1996, ApJ, 470, 706Anders, E., & Grevesse, N. 1989, Geochim. Cosmochim. Acta,53, 197Arnaud, M., Pointecouteau, E., & Pratt, G. W. 2005, A&A, 441,893Bahcall, N. A., & Kulier, A. 2014, MNRAS, 439, 2505Barnes, J. E. 1989, Nature, 338, 123Böhringer, H., Voges, W., Huchra, J. P., et al. 2000, ApJS, 129,435Böhringer, H., Schuecker, P., Guzzo, L., et al. 2004, A&A, 425,367Buote, D. A., Gastaldello, F., Humphrey, P. J., et al. 2007, ApJ,664, 123Cavaliere, A., & Fusco-Femiano, R. 1976, A&A, 49, 137—. 1978, A&A, 70, 677Cravens, T. E. 2000, ApJ, 532, L153Cypriano, E. S., Mendes de Oliveira, C. L., & Sodré, Jr., L. 2006,AJ, 132, 514Dariush, A., Khosroshahi, H. G., Ponman, T. J., et al. 2007,MNRAS, 382, 433Dariush, A. A., Raychaudhury, S., Ponman, T. J., et al. 2010,MNRAS, 405, 1873Desjardins, T. D., Gallagher, S. C., Hornschemeier, A. E., et al.2014, ApJ, 790, 132Doe, S., Nguyen, D., Stawarz, C., et al. 2007, in AstronomicalSociety of the Pacific Conference Series, Vol. 376, AstronomicalData Analysis Software and Systems XVI, ed. R. A. Shaw,F. Hill, & D. J. Bell, 543Donahue, M., Horner, D. J., Cavagnolo, K. W., & Voit, G. M.2006, ApJ, 643, 730 c (cid:13) , 1–17 ossil group origins – VI D’Onghia, E., Sommer-Larsen, J., Romeo, A. D., et al. 2005,ApJ, 630, L109Eckmiller, H. J., Hudson, D. S., & Reiprich, T. H. 2011, A&A,535, A105Eisenstein, D. J., Annis, J., Gunn, J. E., et al. 2001, AJ, 122,2267Ettori, S., Gastaldello, F., Leccardi, A., et al. 2010, A&A, 524,A68Fabian, A. C., & Barcons, X. 1992, ARA&A, 30, 429Fujimoto, R., Mitsuda, K., Mccammon, D., et al. 2007, PASJ,59, 133Girardi, M., Giuricin, G., Mardirossian, F., Mezzetti, M., &Boschin, W. 1998, ApJ, 505, 74Girardi, M., Manzato, P., Mezzetti, M., Giuricin, G., & Limboz,F. 2002, ApJ, 569, 720Girardi, M., & Mezzetti, M. 2001, ApJ, 548, 79Girardi, M., Aguerri, J. A. L., De Grandi, S., et al. 2014, A&A,565, A115 (G14)Harrison, C. D., Miller, C. J., Richards, J. W., et al. 2012, ApJ,752, 12Hess, K. M., Wilcots, E. M., & Hartwick, V. L. 2012, AJ, 144,48Hogg, D. W. 1999, preprint (astro-ph/9905116)Hudson, D. S., Mittal, R., Reiprich, T. H., et al. 2010, A&A,513, A37Ishibashi, W., & Courvoisier, T. J.-L. 2010, A&A, 512, A58Ishisaki, Y., Maeda, Y., Fujimoto, R., et al. 2007, PASJ, 59, 113Jones, L. R., Ponman, T. J., & Forbes, D. A. 2000, MNRAS,312, 139Jones, L. R., Ponman, T. J., Horton, A., et al. 2003, MNRAS,343, 627Kalberla, P. M. W., Burton, W. B., Hartmann, D., et al. 2005,A&A, 440, 775Kawaharada, M., Okabe, N., Umetsu, K., et al. 2010, ApJ, 714,423Kelly, B. C. 2007, ApJ, 665, 1489Khosroshahi, H. G., Gozaliasl, G., Rasmussen, J., et al. 2014,MNRAS, 443, 318Khosroshahi, H. G., Maughan, B. J., Ponman, T. J., & Jones,L. R. 2006, MNRAS, 369, 1211Khosroshahi, H. G., Ponman, T. J., & Jones, L. R. 2007, MN-RAS, 377, 595 (KPJ07)Kuntz, K. D., & Snowden, S. L. 2000, ApJ, 543, 195Kushino, A., Ishisaki, Y., Morita, U., et al. 2002, PASJ, 54, 327Lovisari, L., Reiprich, T. H., & Schellenberger, G. 2015, A&A,573, A118Massaro, E., Giommi, P., Leto, C., et al. 2009, A&A, 495, 691Maughan, B. J., Giles, P. A., Randall, S. W., Jones, C., & For-man, W. R. 2012, MNRAS, 421, 1583Méndez-Abreu, J., Aguerri, J. A. L., Barrena, R., et al. 2012,A&A, 537, A25Miller, E. D., Rykoff, E. S., Dupke, R. A., et al. 2012, ApJ, 747,94Milosavljević, M., Miller, C. J., Furlanetto, S. R., & Cooray, A.2006, ApJ, 637, L9Navarro, J. F., Frenk, C. S., & White, S. D. M. 1997, ApJ, 490,493Osmond, J. P. F., & Ponman, T. J. 2004, MNRAS, 350, 1511Pointecouteau, E., Arnaud, M., & Pratt, G. W. 2005, A&A, 435,1Ponman, T. J., Allan, D. J., Jones, L. R., et al. 1994, Nature,369, 462Ponman, T. J., & Bertram, D. 1993, Nature, 363, 51Popesso, P., Biviano, A., Böhringer, H., & Romaniello, M. 2007,A&A, 464, 451Pratt, G. W., & Arnaud, M. 2005, A&A, 429, 791Pratt, G. W., Croston, J. H., Arnaud, M., & Böhringer, H. 2009,A&A, 498, 361 Proctor, R. N., de Oliveira, C. M., Dupke, R., et al. 2011, MN-RAS, 418, 2054Rasmussen, J., & Ponman, T. J. 2007, MNRAS, 380, 1554Santos, W. A., Mendes de Oliveira, C., & Sodré, Jr., L. 2007,AJ, 134, 1551 (S07)Serlemitsos, P. J., Soong, Y., Chan, K.-W., et al. 2007, PASJ,59, 9Stern, J., & Laor, A. 2012, MNRAS, 423, 600Sugizaki, M., Kamae, T., & Maeda, Y. 2009, PASJ, 61, 55Sun, M., Voit, G. M., Donahue, M., et al. 2009, ApJ, 693, 1142Tawa, N., Hayashida, K., Nagai, M., et al. 2008, PASJ, 60, 11Uchiyama, Y., Maeda, Y., Ebara, M., et al. 2008, PASJ, 60, 35Véron-Cetty, M.-P., & Véron, P. 2010, A&A, 518, A10Vikhlinin, A., Kravtsov, A., Forman, W., et al. 2006, ApJ, 640,691Voevodkin, A., Borozdin, K., Heitmann, K., et al. 2010, ApJ,708, 1376von Benda-Beckmann, A. M., D’Onghia, E., Gottlöber, S., et al.2008, MNRAS, 386, 2345Wu, X.-P., Xue, Y.-J., & Fang, L.-Z. 1999, ApJ, 524, 22Yoshino, T., Mitsuda, K., Yamasaki, N. Y., et al. 2009, PASJ,61, 805Zarattini, S., Barrena, R., Girardi, M., et al. 2014, A&A, 565,A116 (Z14)Zarattini, S., Aguerri, J. A. L., Sánchez-Janssen, R., et al. 2015,A&A, 581, A16Zhang, Y.-Y., Andernach, H., Caretta, C. A., et al. 2011, A&A,526, A105
APPENDIX A: TESTING FGS24 FOR SWCXCONTAMINATION
The NASA WIND-SWE proton flux light curve displays el-evated flux levels greater than × cm − s − during asignificant portion of the FGS24 observation (Fig. A1) whichindicates SWCX photons may contaminate the lower E < keV region of the spectrum (see Section 4.2). To test forevidence of this contamination, we repeat the spectral anal-ysis of Section 6.1 for the time intervals where the flux wasless than × cm − s − . These results are recorded inTable A1 and we find these results are consistent within er-ror with those of using the full timespan of the observation(Table 4). APPENDIX B: CHARACTERIZING THESUZAKU XRT PSF
We determine a radial model for the
Suzaku
XRT PSF tocomplete our image analysis in Section 5.2. Our PSF charac-terization employs archival observations of the X-ray pointsource SS Cyg observed for an effective 52 ks between 2005November 18 and 19 (
Suzaku sequence number 400007010).We clean the SS Cyg event files following the same procedureapplied to our
Suzaku observations (see Section 3).The PSF is characterized using the radial profile of thestacked XIS0+XIS1+XIS3 image of SS Cyg that has beenextracted in the 0.5–10 keV energy range and normalized to1 (Fig. B1). The average PSF full width at half-maximum(FWHM) is found to be ∼ arcsec. Our PSF model con-sists of the sum of two exponentials, as recommended bySugizaki et al. (2009), and thus the model fit to the SS Cyg c (cid:13) , 1–17 A. Kundert et al. T a b l e A . B e s t fi t s p ec t r a l p a r a m e t e r s du r i n ga l o w p r o t o nflu x t i m e i n t e r v a l f o r F G S . F G S r a p , s r c a pe c p o w e r l a w a pe c + p o w e r l a w k T a p ec Z a p ec n o r m a a p ec χ / d . o . f ( χ r ) Γ P L n o r m b P L χ / d . o . f ( χ r ) k T a p ec Z a p ec n o r m a a p ec Γ P L n o r m b P L χ / d . o . f ( χ r ) [ k e V ][ Z ⊙ ][ − ] [ − ] [ k e V ][ Z ⊙ ][ − ] [ − ] . ’ . + . − . . + . − . . + . − . ( . ) . + . − . . + . − . ( . ) . + . − . . . + . − . . . + . − . ( . ) a n o r m a p ec = − π [ D A ( + z ) ] R n e n H d V c m − b n o r m P L h a s un i t s o f ph o t o n s k e V − c m − s − a r c m i n − a t k e V MJD0.00.10.20.30.40.50.60.7 r a t e [ c t s s − ] XIS1 0.4-2 keV
FGS 24
MJD246810 p r o t o n f l u x [ c m − s − ] WIND-SWE
Figure A1.
Top: the observed XIS1 light curve for FGS24. Bot-tom: the WIND-SWE proton flux light curve plotted for the sametime span. Proton flux has been found to be correlated to SWCX.The elevated proton flux levels during the FGS24 observation maypotentially cause significant SWCX contaminating emission.
Table B1.
Best-fitting model to the radial brightness profile ofSS Cyg.Component Parameter Value Unitsexp1 A +0 . − . counts arcsec − c -2.5 +0 . − . − arcsec − r , +115 . − . arcsec exp2 A +2 . − . counts arcsec − c -9.2 +0 . − . − arcsec − r , +38 . − . arcsec background k +0 . − . − counts arcsec − χ / d . o . f( χ r ) brightness profile is: S ( r ) = A e c ( r − r , ) + A e c ( r − r , ) + k, (B1)where the constant k accounts for the background. The best-fitting parameters for this model are recorded in Table B1. APPENDIX C: NOTES ON THE SAMPLE
FGS03 is a Z14 verified fossil system. The AGN (2MASXJ07524421+4556576) associated with the BCG of this sys-tem is both confirmed in the optical (Véron-Cetty & Véron2010) and radio. The radio emission from this object con-sists of strong bipolar jets extending 57 arcsec (Hess et al.2012). This AGN has also been identified as a Type I Seyfert c (cid:13) , 1–17 ossil group origins – VI -4 -3 -2 -1 s u r f a c e b r i g h t n e ss [ c o un t s a r c s e c − ] SS CYG d a t a / m o d e l Figure B1.
Stacked and normalized XIS0+XIS1+XIS3 radialbrightness profile for point-source SS Cyg in the 0.5–10 keV band.The best-fitting model, consisting of the sum of two exponen-tials and a background constant, is plotted in solid blue. Compo-nents of the model are plotted with dashed lines, and residualsare plotted as triangles. Best-fitting parameters for the model arerecorded in Table B1. (Stern & Laor 2012), and appears to dominate the X-rayemission observed from FGS03. The spectrum of this objectis better fit by a power-law ( χ r = 1 . ) than a thermal model( χ r = 1 . ), and no improvement in the fit occurs when athermal component is added to the power-law model. Fur-thermore, our imaging analysis finds a β -model poorly de-scribes the observed surface brightness profile. Z14 find avelocity dispersion of σ v = 259 km s − , the smallest dis-persion of the S07 catalogue. Such a low velocity dispersionis typically associated with a cool ICM temperature, whichwould explain why there appears to be very little thermalemission when compared to a very bright AGN. FGS04 is a fossil candidate and has the coolest mea-sured ICM of our sample ( T X = 2.81 keV). The BCGof this system contains the blazar NVSS J080730+340042(Massaro et al. 2009) and in the radio, Hess et al. (2012)find bipolar jets originating from this source. We do notsee evidence of contribution from this object in the spectralanalysis - the spectrum of FGS04 is fit significantly betterby a thermal model than a power-law (compare a χ r of 1.14to 1.43). FGS09 is a fossil candidate system at z = 0 . .A background z = 0 . AGN (QSO B1040+0110;RA=10:43:03.84, Dec.=+00:54:20.42) is located 15 arcsecfrom the peak X-ray coordinates of FGS09. This AGN isconfirmed in the optical (Véron-Cetty & Véron 2010) andthe radio (Hess et al. 2012) bands. Based on our surfacebrightness profile and spectral analyses, this AGN is signifi-cantly contributing to the observed projected X-ray emissionof FGS09. A large reduced chi-squared of χ r =5.7 is foundfor the β -model fit to the radial brightness profile. And, apower-law model ( χ r = 0 . ) fits the spectrum of FGS09much better than the thermal model ( χ r = 1 . ). FGS14 is a confirmed fossil system and is the largest, hottest, and most X-ray luminous cluster in our sample,with r = 1 Mpc, T X = 5.3 keV, and L X = . × erg s − . Hess et al. (2012) detected radio-loud emission fromtwo central sources; however, we did not see evidence of X-ray bright non-thermal emission in our spectral tests. FGS15 is a rejected fossil candidate (Z14). Thereare a number of contaminating sources in the XISFOV of this source. A radio-loud AGN with an asym-metric jet is associated with the BCG of this system(Hess et al. 2012). Within 40 arcsec of the peak sys-tem X-ray, the background ( z = 0 . ) quasar [VV2010]J114803.2+565411 has been identified optically and in theradio (Véron-Cetty & Véron 2010; Hess et al. 2012). Of thetwo visually distinguishable point sources excluded in ouranalysis, the object closest to the centre of the system is spa-tially consistent with the QSO [VV2010] J114755.9+564948at z = 4 . (Véron-Cetty & Véron 2010). The furthersouth removed point source is located at (RA=11:48:08.38,Dec.=+56:48:18.64). The closest known spatial match tothis object is the radio source NVSS J114838+565327 lo-cated ∼ β -model ( χ r =5.2) poorly fits the observedemission, and additionally the best-fitting spectral model ofFGS15 is a power-law. For this object, it is possible multipleAGN are contributing to the observed emission; however, asnoted by Z14, FGS15 could also be a filament due to itssmall number of constituent galaxies with large differencesin velocity. FGS24 is a rejected fossil candidate. No associatedAGN were identified in the literature. However, the spec-trum of FGS24 is better fit by a power-law than a thermalmodel (compare a χ r of 1.33 to 1.38). FGS24 was observedduring a period of potentially strong SWCX emission. Whilewe found the best-fitting spectral parameters of the full ob-servation match those of the isolated time interval of lowproton flux, it is possible SWCX contamination is occur-ring even during this interval, obscuring the emission fromFGS24. FGS25 is a non-fossil galaxy cluster (Z14). It is thesecond hottest cluster in our sample with T X = 3.92 keV anda corresponding estimated mass of M = 2 . × M ⊙ .Hess et al. (2012) find a radio-loud central point source inthis cluster; however, our spectral analysis indicates no pointsource contribution as the FGS25 spectrum is much betterdescribed by a thermal model ( χ r = 0 . ) than a power-lawmodel ( χ r = 1 . ). FGS26 is a Z14 confirmed fossil with T X = 3.3 keV and L X = . × erg s − . We find no associated significantnon-thermal signatures in the spectrum. FGS27 is a confirmed fossil with measured global prop-erties of T X = 3.3 keV and L X = . × erg s − . Ourspectral analysis does not indicate contribution of significantnon-thermal emission. FGS30 is a confirmed fossil with measured global prop-erties of T X = 3.4 keV and L X = . × erg s − . A radio-loud AGN (2MASX J17181198+5639563) is associated withits bright central galaxy (Hess et al. 2012). The spectrum ofFGS30 is better described by the thermal model ( χ r = 1 . )in comparison to the power-law model ( χ r = 1 . ). c (cid:13)000