How typical is the Coma cluster?
aa r X i v : . [ a s t r o - ph . C O ] D ec Mon. Not. R. Astron. Soc. , 000–000 (0000) Printed 20 February 2018 (MN L A TEX style file v2.2)
How typical is the Coma cluster?
Kevin A. Pimbblet , , , ⋆ , Samantha J. Penny , , Roger L. Davies Department of Physics and Mathematics, University of Hull, Cottingham Road, Hull, HU6 7RX, UK School of Physics, Monash University, Clayton, Victoria 3800, Australia Monash Centre for Astrophysics (MoCA), Monash University, Clayton, Victoria 3800, Australia Department of Physics, University of Oxford, Denys Wilkinson Building, Keble Road, Oxford OX1 3RH, UK
Accepted ... ; Received ... ; in original ...
ABSTRACT
Coma is frequently used as the archetype z ∼ ∼ σ ) X-ray temperature set against clusters of comparable mass.By de-redshifting our control sample cluster galaxies star-formation rates using a fitto the galaxy main sequence evolution at z < .
1, we determine that the typicalstar-formation rate of Coma galaxies as a function of mass is higher than for galaxiesin our control sample at a confidence level of >
99 per cent. One way to alleviatethis discrepency and bring Coma in-line with the control sample would be to havethe distance to Coma to be slightly lower, perhaps through a non-negligible peculiarvelocity with respect to the Hubble expansion, but we do not regard this as likelygiven precision measurements using a variety of approaches. Therefore in summary,we urge caution in using Coma as a z ∼ Key words: galaxies: clusters: individual: Coma cluster — galaxies: clusters: general— galaxies: evolution — X-rays: galaxies: clusters
Clusters of galaxies span a wide range of physical condi-tions and internal configurations. At the high mass end oftheir mass distribution ( ∼ solar masses), clusters maycontain several thousand member galaxies that are orbitingwith velocity dispersions over 1000 kms − (cf. Pimbblet etal. 2006; Ebeling et al. 2010). They are also rare celestial ob-jects: they form from the gravitational collapse of extremelylarge perturbations within the primorial density field (e.g.Zel’Dovich 1970; Doroshkevich & Shandarin 1978) and con-tinue to grow at all epochs through the accretion of freshmaterial; a large fraction of galaxies being funnelled directlyto them through the filaments of the cosmic web (Pimbblet,Drinkwater & Hawkrigg 2004). From the point of view ofstudying galaxy evolution, clusters of galaxies offer excellenttest-beds as they contain a range conditions from their out-skirts (which may contain filaments and under-dense ‘void’regions that galaxies are being accreted from) through to ⋆ email: [email protected] high density cores that contain a dense, hot (10 –10 K)X-ray emitting gas that is capable of stripping an infallinggalaxy of its own star-forming gas (Gunn & Gott 1972; Cay-atte et al. 1990; Quilis et al. 2000; Boselli & Gavazzi 2006).Indeed, galaxies that are located at the centre of clusters (orhigh density regions of the Universe) have long been notedto possess systematically different properties (star-formationrates; colours; morphologies; masses) to those in low densityregions (e.g. Dressler 1980; Lewis et al. 2002; G´omez et al.2003; Baldry et al. 2006; Bamford et al. 2009; Wilman &Erwin 2012; amongst many others). To address questionsconcerning the evolution of galaxies within these structures,samples of self-similar structures (and/or their likely pro-genitors) need to be assembled across cosmic time.The Coma cluster (also known as Abell 1656 in thecatalogue of Abell 1958) is the closest galaxy cluster of itsmass (recently derived to be 1 . × solar masses througha weak lensing analysis by Kubo et al. 2007) to us. Thishas lead to Coma being extensively used as a redshift z ≈ c (cid:13) K. A. Pimbblet, S. J. Penny, R. L. Davies
Dickinson 1995; Smith, Driver & Phillipps 1997; Kodamaet al. 1998; Jørgensen et al. 1999; Jones, Smail & Couch2000; Kodama & Bower 2001; van Dokkum et al. 2001; LaBarbera et al. 2002; Rusin et al. 2003; De Lucia et al. 2004;Ellis & Jones 2004; Poggianti et al. 2004; Fritz et al. 2005;Holden et al. 2005 De Lucia et al. 2007; Moran et al. 2007;van Dokkum & van der Marel 2007; D’Onofrio et al. 2008;Giard et al. 2008; Ascaso et al. 2009; Bai et al. 2009; Lah etal. 2009; Stott et al. 2009).Yet it is not clear how representative (or ‘typical’) theComa cluster is for clusters of its mass. To illustrate thispoint, we note two recent examples. Stott et al. (2009)present an analysis of how the slope of the colour-magnituderelation (Visvanathan & Sandage 1977) of clusters varieswith redshift. They find that the rest-frame slope evolvesaccording to (1 + z ) . (see their Fig. 7). Yet, the slopefor the Coma cluster lies at least 2 σ away (steeper) fromthis relationship and its absolute value is much more in-linewith what might be expected of a z ∼ . z < .
08. Stott et al. (2009) attribute thismildly unusual slope to a lower than average dwarf-to-giantratio along its red sequence (Stott et al. 2007) that suggestsit is still undergoing significant faint end evolution. WhilstStott et al.’s result is likely not a statistically significant is-sue, other studies yield stronger issues with the use of Comaas a z ≈ . < z < .
25 and compare themto Coma (using data from Aguerri et al. 2004). They findthat the scales of the discs of late-type galaxies in the highredshift clusters are significantly different to Coma. They of-fer two conclusions: either spiral galaxies have undergone aremarkable and very strong evolution over the past 2.5 Gyr,or ‘Coma is in some way anomalous’ (Ascaso et al. 2009).Much earlier studies that concentrate on Coma itself de-scribe the cluster as ‘rich’, ‘regular’ and (or) ‘relaxed’ (e.g.Kent & Gunn 1982 retain the assumption of the cluster be-ing in equilibrium; see also Noonan 1961; Omer, Page & Wil-son 1965 and references therein). Evidence subsequently ac-cumulated that Coma was anything but a local archetype forrelaxed and regular clusters: Henriksen & Mushotzky (1986)used X-ray observations to invalidate the assumption of anisothermal sphere (see also Johnson et al. 1979; Briel, Henry& Boehringer 1992; White, Briel & Henry 1993; Vikhlinin,Forman & Jones 1997; Neumann et al. 2003); the cluster con-tains multiple D class galaxies (Beers & Geller 1983); andimportantly the velocity distribution of the galaxy membersthemselves revealed substructure (Fitchett & Webster 1987;Merritt 1987; Mellier et al. 1988; Colless & Dunn 1996; Gam-bera et al. 1997; Edwards et al. 2002; Adami et al. 2009; seealso Conselice & Gallagher 1998).The central thesis of this work is to present a novelinvestigation of how typical the Coma cluster is in threewell-defined and distinct ways that are well-used in the lit-erature. This comprises: (i) an investigation in to the X-ray properties (particularly temperature and luminosity) ofComa in comparison to analogue clusters; (ii) a consider-ation of how kinematically perturbed or relaxed analogue clusters are to Coma; (iii) a determination of how ‘active’ –in the sense of star-formation – the galaxies that make upanalogous clusters are compared to Coma. The format forthis work is as follows. In Section 2, we describe the cre-ation of a set of control clusters that are analogous to Comain mass from available SDSS and X-ray data. We examinethe X-ray properties of Coma in comparison to the controlsample and an extended sample in Section 3. Section 4 dealswith the kinematics of the galaxies contained in the clustersand in Section 5, we examine the star-formation rates of theconstituent galaxies in Coma and the control sample. Ourresults are summarized in Section 6. Throughout this work,we have used the Spergel et al. (2007) standard, flat cosmol-ogy in which Ω M = 0 . Λ = 0 .
762 and H = 73 km s − Mpc − . L X = 7 . × erg s − measured in the0.1–2.4 keV band (Reiprich & B¨ohringer 2002). This levelof emission is comparable with some of the most massiveclusters in the Universe (cf. Ebeling et al. 2001; Pimbbletet al. 2001). We therefore would like to select clusters withcomparable L X in the 0.1–2.4 keV band, but balance thiswith a need to have a sufficiently large control sample tocontrast Coma against. We therefore select clusters within5 × ergs − of Coma’s X-ray luminosity. Since X-rayluminosity can predict cluster mass with an accuracy of > L X selection (Popesso etal. 2005). Secondly, we would like to select galaxy clustersto be at a comparable stage in their evolution as Coma.We firstly note that Kodama & Smail (2001) suggest thetime-scale for galaxy morphological transformation withinclusters may be as short as 1 Gyr if gas starvation effects arestrong (see also Bekki, Couch & Shioya 2002; Moran et al.2006; Tonnesen et al. 2007; Boselli et al. 2008). Therefore wewish to select clusters within a < z ∼ .
08 to selectour clusters from.We use the Base de Donnees Amas de Galaxies X(BAX) X-Ray Clusters Database (Sadat et al. 2004) to se-lect clusters from using the above criteria. This yields a to-tal of 47 clusters. Of these, one is Coma and a further 13(30 per cent) are within the spatial limits of SDSS – thiscriteria of being within the observational bounds of SDSS isonly applied after the X-ray selection within BAX. We detailthe global properties of these clusters in Table 1, alongsideComa. We note that the clusters in the control sample have c (cid:13) , 000–000 ow typical is the Coma cluster? Table 1.
The sample of clusters used in this work. The coordinates specify the Vizier position of the cluster. The X-ray luminositiesin the 0.1–2.4 keV band ( L X ) and temperatures ( T X ) are sourced from BAX (Sadat et al. 2004) which is a compilation of X-ray dataderiving from many diverse literature sources. We cite the sources of these values below the table using brackets next to each value. Thevirial radius ( R virial ) is computed from σ cz ; see text for details. Bautz-Morgan (B-M) types have been sourced from NED except forthe Zwicky clusters which we have determined ourselves.Name RA Dec B-M L X T X cz σ cz Adopted(J2000) (J2000) type ( × erg s − ) (keV) (km s − ) (km s − ) R (Mpc)Abell 85 00 41 38 −
09 20 33 I 9.41 (i) 6.45 +0 . − . (a) 16537 ±
58 898 +45 − −
01 15 47 II-III 3.30 (i) 5.62 +0 . − . (b) 13381 ±
56 917 +43 − ±
109 681 +94 − +0 . − . (a) 23109 ±
54 867 +41 − +0 . − . (c) 21593 ±
113 1637 +86 − +0 . − . (a) 18773 ±
55 719 +42 − +0 . − . (d) 22915 ±
80 1018 +61 − +0 . − . (c) 13408 ±
88 1064 +68 − +0 . − . (d) 23083 ±
52 841 +39 − +0 . − . (a) 22213 ±
100 1887 +75 − +0 . − . (a) 9176 ±
44 761 +33 − +0 . − . (c) 24071 ±
82 1223 +62 − +0 . − . (e) 7166 ±
54 1639 +40 − a mean L X = 5 . ± . × erg s − – only ∼ σ less thanComa’s.From this sample, we exclude NRGB045 on the groundsthat it has an anomolously low T X value (0.83 keV). Thisis due the NRGB045 being more akin to a group than acluster. Indeed, recent work by Stott et al. (2012) suggeststhat any galaxy grouping with T X < cz ) and velocity dispersion( σ cz ) from the ‘gapping’ technique of Zabludoff, Huchra &Geller (1990; 1993) which iteratively eliminates any galaxyfrom the computation of cz that is deviant by more than3 σ cz from cz . Errors on σ cz are generated following Denese,de Zotti, G. & di Tullio (1980). Although this method sam-ples a factor of ∼ σ cz of cz . An analogous approach is taken for Coma, butusing a 2 degree radius (a 3.4 Mpc radius). To place theclusters on to a common, physically meaningful scale, welimit our subsequent analysis to those galaxies to within r ≈ R V irial = 0 . σ cz (Girardi et al. 1998), where r is the clustocentric radius at which the mean interior densityis 200 times the critical density; this value is well approxi- mated by R virial . Although we could compute this radii inother ways (e.g. Carlberg et al. 1997), we emphasize thatthis approximation is sufficient to serve to place our clus-ters on to a common scale. These values are tabulated inTable 1. Although it is known that there is considerablescatter in the L X - σ cz relationship (Popesso et al. 2005), thefirst conclusion to be drawn here is that Coma’s velocitydispersion is not atypical compared to the control sample(which has a mean of 1043 ±
372 kms − ), but is is one of thelargest given how we have selected the galaxy members. Wepoint out that the control sample has a full range of Bautz-Morgan (1970) classifications (Table 1) – meaning we covera full range of galaxy cluster configurations and morpholo-gies, ranging from those with obvious cD galaxies centrallylocated in the clusters those lacking such a galaxy in entirety.Coma as a type II cluster that has two obvious, brightestcluster galaxies is not atypical against this control sample:we do not regard it as more dynamically evolved than thecontrol sample. From Table 1, it is already clear that Coma has the largestX-ray temperature (8.25 keV) out of all the comparable clus-ters selected within SDSS. The mean temperature of ourcontrol sample is T X = 5 . ± . . σ lower thanthe temperature of Coma. Such a large temperature meansthat the physical conditions inside Coma may actively reg-ulate the star-formation of galaxies contained therein. Forexample, Urquhart et al. (2010) notes that high T X clusters c (cid:13) , 000–000 K. A. Pimbblet, S. J. Penny, R. L. Davies
Figure 1.
X-ray temperatures for clusters extracted from BAX within 1 × ergs − of Coma’s X-ray luminosity that have T X valuesavailable. Clusters below z = 0 . T X of our control sample (i.e. those clusters inside the SDSS boundary that are within 5 × ergs − of Coma’s X-ray luminosity) is denoted by the solid horizontal line and a few standard deviations either side of this is represented bythe dotted lines, as labelled. Coma has one of the largest T X values for this narrow L X range and is ∼ . σ above the control sample’smean T X value. have a much lower fraction of photometrically blue galax-ies (i.e. Butcher-Oemler fraction; Butcher & Oemler 1984)than low T X clusters and are highly unlikely to contain anyextremely blue galaxies. Further, Popesso et al. (2007b) andAguerri et al. (2007) find an anti-correlation between L X and cluster blue fraction which supports this finding giventhe scaling between L X and T X . This is reflected in thework of Poggianti et al. (2006) who demonstrate a broadanti-correlation between cluster velocity dispersion (a pa-rameter that also scales with L X ; Dav´e et al. 2002) and thefraction of star-forming cluster galaxies. Popesso et al. (2005) report the scaling relationship be-tween L X and T X in detail and show that there is both atrend and appreciable scatter between the two variables (seealso Dav´e et al. 2002). Although Coma’s T X value may besignificantly larger than our control sample, we have useda factor of 2 range in L X to draw this conclusion from. Todetermine if its T X is truly anomolously high, we need toselect clusters in a much narrower range of L X . We turnagain to BAX to do this and select all available clusterswithin 1 × ergs − of Coma’s X-ray luminosity that alsohave a reliable X-ray temperature measurement available. c (cid:13) , 000–000 ow typical is the Coma cluster? In Fig. 1 we plot this narrow range of L X against T X forall available clusters. Coma is again seen to have one of thehighest temperatures for all clusters in this range – bothabove and below the redshift cut-off of our control sampleof z = 0 . σ of the mean T X of this narrower L X range sample. That said, there is onlyone cluster either side of this redshift that has a larger X-raytemperature .We therefore conclude that Coma’s X-ray temperatureis comparatively high: both against our control sample, andagainst all available clusters in a much narrower L X range. In this section, we address the second of our comparisons ofComa to the control sample using global cluster kinematicalapproaches. Depending on cosmological parameters such asthe matter density of the Universe, it might be expected thata rich cluster of galaxies (i.e. such as the ones that are in oursample) have perhaps had as much as half of their mass ac-creted within the past ∼ few Gyr (e.g. Lacey & Cole 1993).Under such circumstance, it can be expected that a largefraction of rich clusters exhibit measurable sub-clustering.Coma is already well-known to possess sub-clustering (seeabove). But, what fraction of our control sample also ex-hibits sub-clustering? To determine this fraction, we use theapproach of Dressler & Shectman (1988; DS) to evaluateif the clusters possess significant sub-clustering. The test ispowerful: Pinkney et al. (1996) report that the DS approachis the most sensitive test for sub-clustering from a swatheof tests that they evaluated. The method works by comput-ing a local mean local velocity ( cz local ) and local velocitystandard deviation ( σ local ) of a galaxy and its ten nearestneighbours. These values are subsequently compared to theparent cluster’s mean velocity and σ z such that: δ = (cid:16) N local + 1 σ v (cid:17) [( cz local − cz ) + ( σ local − σ v ) ] (1)where δ is a measure of the deviation of the individualgalaxy. The parameter of merit, ∆, is computed as the sum-mation of all δ terms. This is contrasted to a Monte Carlore-simulation of the cluster where the galaxy velocities havebeen randomly shuffled to each galaxy to generate P (∆)and thereby estimate the confidence level that the clustercontains sub-structure.Before we apply the DS test to our control sample, weneed to not only use the cluster membership criteria de-rived above and limit the members to within R V irial , butalso limit the cluster members to a similar absolute lumi-nosity range and mass range. This is necessary since sub-structure is strongly dependant on the galaxy luminosityrange considered (Aguerri & Sanchez-Janssen 2010). Thisis achieved by considering the highest redshift cluster in thecontrol group: Abell 2255. For this cluster, the SDSS limiting We re-affirm the note made by Valtchanov et al. (2002) aboutAbell 1451 ( T X = 13 . L X - T X scaling relation(e.g., Popesso et al. 2005). This cluster merits future follow-upto discern the impact and potential cause of such an extremetemperature. Figure 2.
Mass and absolute luminosity of all galaxies in the con-trol sample (smaller, black dots) with the contribution from themost distant cluster in our sample, Abell 2255, overlaid (larger,red dots). The SDSS limiting apparent magnitude of r = 17 . . apparent magnitude of r = 17 .
77 corresponds to an absolutevalue of − .
85 (Fig. 2). At this limit, we are mass completeto log(stellar mass)= 10 . P (∆) statistic that is < . R V irial ).We note that this remains constant even if we ignore theabsolute magnitude limit imposed above. Given the com-paratively large velocity dispersion of these clusters, this isperhaps expected (Hou et al. 2012). Moreover, from ΛCDMsimulations of clusters, Knebe & Muller (2000) demonstratethat some 30 per cent of all clusters should exhibit subclus-tering due to inter-cluster merger and infall activity (moduloa slightly different selection criteria). We therefore regardComa (and, indeed, our control group) as being ”typical”for clusters in a ΛCDM Universe for the level of substruc-ture observed at our limits. c (cid:13) , 000–000 K. A. Pimbblet, S. J. Penny, R. L. Davies
Figure 3.
Velocity dispersion profiles of our clusters. For each cluster, σ P ( R ) is normalized to the central value. ZwCl1215 (top right)is arguably the only cluster to display a strongly rising profile with radius whereas the other clusters either have a flat, falling, orcombination profile. In recent years, a number of authors have probed how thevelocity dispersion profile of clusters is affected by vari-ous cluster-intrinsic factors such as substructure (Hou et al.2012) as well as potentially the dwarf to giant ratio (Pimb-blet & Jensen 2012) and the occupancy of the cluster bydifferent spectral classes of galaxy (Rood et al. 1972). Tocomplement the above analysis, we now compute the veloc-ity dispersion profile ( σ P ( R )) of each of our clusters follow-ing the prescription of Bergond et al. (2006; see also Hou etal. 2012). Formally, σ P ( R ) = s P i w i ( R )( x i − x ) P i w i ( R ) (2)where x i are the measured radial velocities of each galaxyand x is the mean recession velocity of the cluster taken fromTable 1. The weighting factors, w i , are applied such that: w i ( R ) = 1 σ R exp (cid:18) ( R − R i ) σ R (cid:19) (3)where σ R , the kernel width, is a free parameter that wearbitrarily set to 0 . R V irial . The velocity dispersion pro- files computed in this manner are displayed in Fig. 3. Inter-estingly, the clusters with significant sub-structure are notseen to have a rising velocity dispersion profile. This arguesthat any local kinematic group of galaxies may be at a latestage of homogenization with the wider cluster. This is incontrast to ZwCl 1215+0400 which does have a markedlyrising profile and lack obvious substructure. This may becaused by multiple sub-clumps at large radii infalling forthe first time. In comparison, Coma is quite un-remarkableset against these profiles.
In this section, we determine the star-formation activity lev-els for cluster members in Coma and the control sample.One way in which to do this is to use the galaxy mainsequence (Noeske et al. 2007; and references therein): a plotof star formation rate against galaxy mass. This sequence isknown to evolve with redshift – at high z, the average starformation rate of galaxies is higher per galaxy mass than atlower z; the evolution in the trend being largely attributedto gas exhaustion. Therefore, if we are to use the galaxymain sequence to probe the activity levels in Coma and the c (cid:13) , 000–000 ow typical is the Coma cluster? control sample, we must first correct for this redshift evolu-tion. We accomplish this by accessing all SDSS galaxies in0.005 redshift bins up to z = 0 . of log(stellarmass)=10.4–10.6 galaxies (the choice of this mass range isarbitrary, but is sufficiently representative of our own sampleand balances the needs to have good statistics to computethe redshift evolution of the main sequence from). The re-sults of this are displayed in Fig. 4. We fit the data with alinear relationship which has a gradient of 7 . ± .
21 in thisrange. Although the actual evolutionary relation will likelyby a higher order of (1 + z ), this linear relation is sufficientto describe these data at z < . < . −
11 to differentiate be-tween active and passive galaxies. In Fig. 6 we plot thespecifc star-formation rate of Coma galaxies and the de-redshifted control sample as a function of galaxy mass. Thefraction of galaxies that are active by this definition are0 . ± .
02 in Coma versus 0 . ± .
02 for the control sam-ple. This is ∼ σ (depending on rounding) difference be-tween the two samples. This appears to support (albeit at aweaker level) the inference of the two dimensional KS test:the galaxies in Coma are systematically different to the con-trol sample. At the outset, we aimed to investigate three facets of Coma’sgalaxies in comparison to a control sample: an X-ray temper-ature and luminosity comparison, a kinematic comparison,and a star formation activity comparison. One area that wehave deliberately avoided is an examination of the luminos-ity function of Coma. This is on the grounds that it has Star-formation rates for the galaxies are sourced from the SDSSvalue-added catalogue which are computed as per Brinchmann etal. (2004) using model fitting.
Figure 4.
Evolution of the galaxy main sequence for log(stellarmass)=10.4–10.6 SDSS galaxies up to z = 0 .
1. The points arethe median star-formation rates per redshift bin, whilst the solidlines give the interquartile range of the distribution. As notedby Noeske et al. (2007), the range of star-formation also evolveswith z . The linear fit (dotted line) to the data has a gradient of7 . ± .
21 and we use this fit to evolve all the data in the controlsample to Coma’s redshift with in the subsequent analysis. already been well-studied in comparison to other clusters atmultiple wavelengths (recent examples include but are notlimited to: Yamanoi et al. 2012; Bai et al. 2009; Corteseet al. 2008; Adami et al. 2007) and will likely follow thekinematic results for the control sample (in the sense thatmultiple components may be reflected in a superposition offunctions; see also Tempel et al. 2009).One way in which the situation of higher T X coupledwith higher star-formation rate per galaxy mass bin beinglarger in Coma compared to the control sample might bearranged is if the bluer galaxies in Coma are just arrivingin to the cluster environment (given that a hotter intra-cluster medium should inhibit galaxy star formation subse-quently). This ties with Mahajan et al. (2012)’s finding: ahigh galaxy density in the infalling and filamentary regionsof clusters such as Coma inevitably leads to a greater rate ofgalaxy-galaxy interaction and consequentially an increasedstarburst rate. But the problem with this interpretation isthat there are ∼ equally massive clusters in the control sam-ple by design (i.e. the X-ray selection used here).There have been hints in the literature that some of thespecial features of Coma might be allieviated if the distanceto Coma was slightly lower. Consider for example Fig. 6 ofvan Dokkum & van der Marel (2007) which shows that themass to light ratio of Coma is similar to that of z ∼ . c (cid:13) , 000–000 K. A. Pimbblet, S. J. Penny, R. L. Davies peculiar velocity with respect to the Hubble flow (e.g., to-ward the Shapley concentration). An interesting facet of thishypothetical change would be a driving of the star-formationrates of Coma galaxies lower – bringing them more in linewith the de-redshifted control sample points. Given resultsthat suggest Coma has been reported to have negligible pe-culiar velocity (e.g., Bernardi et al. 2002) and a variety ofmeasurements agreeing within uncertainty on its distance(e.g., Capaccioli et al. 1990; D’Onofrio et al. 1997; Jensen etal. 1999; Kavelaars et al. 2000; Liu & Graham 2001), we donot view this as a likely scenario; we supply it simply as anillustration.In summary, in this work we have shown:(i) Although Coma has a large velocity dispersion, itis not atypical for a cluster of its L X . However, the X-raytemperature of Coma is rather high: some 2 . σ hotter thanour control sample. Even considering all clusters availablewith a published T X within 1 × ergs − of Coma revealsit has one of the highest temperatures for all clusters in therange. Given the relationship between T X and cluster galaxyproperties we urge strong caution in using Coma as a z ∼ R V irial . Comais therefore un-remarkable in this regard.(iii) The velocity dispersion profiles of the control sam-ple contain a mixture of rising, falling, flat and combinationprofiles. Coma is un-remarkable set against this backgroundand reinforces the above conclusion that Coma is kinemati-cally normative for clusters of its ilk.(iv) The general star-formation rate of Coma clustergalaxies inferred from the galaxy main sequence is system-atically higher than for the control sample. A two dimen-sional KS test rejects the hypothesis that the two samplesare drawn from the same parent population with over 99per cent confidence. Further, the fraction of actively starforming galaxies by the definition of McGee et al. (2011)is 0 . ± .
02 for Coma, versus 0 . ± .
02 for the controlsample. We note in speculation that this discrepency couldbe alleviated if the distance to Coma were smaller.Thus, whilst Coma might be kinematially “typical”,the galaxies contained within are less suppressed in star-formation rate than the comparison clusters. We consequen-tially urge caution in using Coma as a z ∼ ACKNOWLEDGEMENTS
KAP thanks Christ Church College, Oxford, for their hospi-tality whilst the bulk of this work was being undertaken. SJPis a Super Science Fellow at Monash University. KAP andSJP thank the Australian Research Council for their sup-port through grant number FS110200047. We thank RyanHoughton and Martin Bureau for valuable discussion duringthe preparation of this work.We would like to express our gratitude to the anony-
Figure 5.
Galaxy main sequence for Coma (red triangles) ver-sus the de-redshifted control sample (black dots). The galaxies inComa have a higher systematic average star formation rate at agiven stellar mass than the control sample.
Figure 6.
Specific star-formation rates as a function of galaxymass. Symbols are the same as per Fig. 5. The horizontal linedenotes the McGee et al. (2011) delimiter between active (abovethe line) and passive (below the line) galaxies. The fraction ofactive galaxies differ between the two samples at a ∼ σ level.c (cid:13) , 000–000 ow typical is the Coma cluster? REFERENCES
Abazajian K. N., et al., 2009, ApJS, 182, 543Abell G. O., 1958, ApJS, 3, 211Adami C., et al., 2009, A&A, 507, 1225Adami C., Durret F., Mazure A., Pell´o R., Picat J. P., West M.,Meneux B., 2007, A&A, 462, 411Aguerri J. A. L., Iglesias-Paramo J., Vilchez J. M., Mu˜noz-Tu˜n´onC., 2004, AJ, 127, 1344Aguerri J. A. L., S´anchez-Janssen R., Mu˜noz-Tu˜n´on C., 2007,A&A, 471, 17Aguerri J. A. L., S´anchez-Janssen R., 2010, A&A, 521, A28Arnaud M., et al., 2001, A&A, 365, L67Ascaso B., Aguerri J. A. L., Moles M., S´anchez-Janssen R., Bet-toni D., 2009, A&A, 506, 1071Ascaso B., Moles M., Aguerri J. A. L., S´anchez-Janssen R., VarelaJ., 2008, A&A, 487, 453Bai L., Rieke G. H., Rieke M. J., Christlein D., Zabludoff A. I.,2009, ApJ, 693, 1840Bahcall N. A., 1972, AJ, 77, 550Baldry I. K., Balogh M. L., Bower R. G., Glazebrook K., NicholR. C., Bamford S. P., Budavari T., 2006, MNRAS, 373, 469Bamford S. P., et al., 2009, MNRAS, 393, 1324Bautz L. P., Morgan W. W., 1970, ApJ, 162, L149Beers T. C., Geller M. J., 1983, ApJ, 274, 491Bekki K., Couch W. J., Shioya Y., 2002, ApJ, 577, 651 Bergond G., Zepf S. E., Romanowsky A. J., Sharples R. M.,Rhode K. L., 2006, A&A, 448, 155Bernardi M., Alonso M. V., da Costa L. N., Willmer C. N. A.,Wegner G., Pellegrini P. S., Rit´e C., Maia M. A. G., 2002,AJ, 123, 2159Boselli A., Gavazzi G., 2006, PASP, 118, 517Boselli A., Boissier S., Cortese L., Gavazzi G., 2008, ApJ, 674,742Briel U. G., Henry J. P., Boehringer H., 1992, A&A, 259, L31Brinchmann J., Charlot S., White S. D. M., Tremonti C., Kauff-mann G., Heckman T., Brinkmann J., 2004, MNRAS, 351,1151Butcher H., Oemler A., Jr., 1984, ApJ, 285, 426Capaccioli M., Cappellaro E., della Valle M., D’Onofrio M.,Rosino L., Turatto M., 1990, ApJ, 350, 110Carlberg R. G., et al., 1997, ApJ, 476, L7Cayatte V., van Gorkom J. H., Balkowski C., Kotanyi C., 1990,AJ, 100, 604Colless M., Dunn A. M., 1996, ApJ, 458, 435Cortese L., Gavazzi G., Boselli A., 2008, MNRAS, 390, 1282Conselice C. J., Gallagher J. S., III, 1998, MNRAS, 297, L34De Lucia G., et al., 2004, ApJ, 610, L77De Lucia G., et al., 2007, MNRAS, 374, 809Danese, L., de Zotti, G., di Tullio, G., 1980, A&A, 82, 322Dav´e R., Katz N., Weinberg D. H., 2002, ApJ, 579, 23Dressler A., 1980, ApJ, 236, 351Dressler A., Shectman S. A., 1988, AJ, 95, 985D’Onofrio M., Capaccioli M., Zaggia S. R., Caon N., 1997, MN-RAS, 289, 847D’Onofrio M., et al., 2008, ApJ, 685, 875Doroshkevich A. G., Shandarin S. F., 1978, MNRAS, 182, 27Ebeling H., Edge A. C., Henry J. P., 2001, ApJ, 553, 668Ebeling H., Edge A. C., Mantz A., Barrett E., Henry J. P., MaC. J., van Speybroeck L., 2010, MNRAS, 407, 83Edwards S. A., Colless M., Bridges T. J., Carter D., MobasherB., Poggianti B. M., 2002, ApJ, 567, 178Ellis S. C., Jones L. R., 2004, MNRAS, 348, 165Fasano G., Franceschini A., 1987, MNRAS, 225, 155Fitchett M., Webster R., 1987, ApJ, 317, 653Fritz A., Ziegler B. L., Bower R. G., Smail I., Davies R. L., 2005,MNRAS, 358, 233Gambera M., Pagliaro A., Antonuccio-Delogu V., Becciani U.,1997, ApJ, 488, 136Giard M., Montier L., Pointecouteau E., Simmat E., 2008, A&A,490, 547Girardi M., Giuricin G., Mardirossian F., Mezzetti M., BoschinW., 1998, ApJ, 505, 74G´omez P. L., et al., 2003, ApJ, 584, 210Gunn J. E., Gott J. R., III, 1972, ApJ, 176, 1Henriksen M. J., Mushotzky R. F., 1986, ApJ, 302, 287Henry J. P., 2004, ApJ, 609, 603Holden B. P., et al., 2005, ApJ, 626, 809Hou A., et al., 2012, MNRAS, 421, 3594Ikebe Y., Reiprich T. H., B¨ohringer H., Tanaka Y., Kitayama T.,2002, A&A, 383, 773Jensen J. B., Tonry J. L., Luppino G. A., 1999, ApJ, 510, 71Johnson M. W., Cruddace R. G., Fritz G., Shulman S., FriedmanH., 1979, ApJ, 231, L45Jones L., Smail I., Couch W. J., 2000, ApJ, 528, 118Jørgensen I., Franx M., Hjorth J., van Dokkum P. G., 1999, MN-RAS, 308, 833Kavelaars J. J., Harris W. E., Hanes D. A., Hesser J. E., PritchetC. J., 2000, ApJ, 533, 125Kent S. M., Gunn J. E., 1982, AJ, 87, 945Kodama T., Arimoto N., Barger A. J., Arag’on-Salamanca A.,1998, A&A, 334, 99Kodama T., Bower R. G., 2001, MNRAS, 321, 18Kodama T., Smail I., 2001, MNRAS, 326, 637c (cid:13) , 000–000 K. A. Pimbblet, S. J. Penny, R. L. Davies
Knebe A., M¨uller V., 2000, A&A, 354, 761Kubo J. M., Stebbins A., Annis J., Dell’Antonio I. P., Lin H.,Khiabanian H., Frieman J. A., 2007, ApJ, 671, 1466La Barbera F., Busarello G., Merluzzi P., Massarotti M., Capac-cioli M., 2002, ApJ, 571, 790Lacey C., Cole S., 1993, MNRAS, 262, 627Lah P., et al., 2009, MNRAS, 399, 1447Lewis I., et al., 2002, MNRAS, 334, 673Liu M. C., Graham J. R., 2001, ApJ, 557, L31Mahajan S., Raychaudhury S., Pimbblet K. A., 2012, MNRAS,427, 1252Marini F., et al., 2004, MNRAS, 353, 1219Matsumoto H., Pierre M., Tsuru T. G., Davis D. S., 2001, A&A,374, 28McGee S. L., Balogh M. L., Wilman D. J., Bower R. G., MulchaeyJ. S., Parker L. C., Oemler A., 2011, MNRAS, 413, 996Mellier Y., Soucail G., Fort B., Mathez G., 1988, A&A, 199, 13Merritt D., 1987, ApJ, 313, 121Moran S. M., Loh B. L., Ellis R. S., Treu T., Bundy K.,MacArthur L. A., 2007, ApJ, 665, 1067Moran S. M., Ellis R. S., Treu T., Salim S., Rich R. M., SmithG. P., Kneib J.-P., 2006, ApJ, 641, L97Neumann D. M., Lumb D. H., Pratt G. W., Briel U. G., 2003,A&A, 400, 811Noeske K. G., et al., 2007, ApJ, 660, L43Noonan T. W., 1961, Ph.D. Thesis, California Institute of Tech-nologyOmer G. C., Jr., Page T. L., Wilson A. G., 1965, AJ, 70, 440Peacock J. A., 1983, MNRAS, 202, 615Pimbblet K. A., Smail I., Edge A. C., Couch W. J., O’Hely E.,Zabludoff A. I., 2001, MNRAS, 327, 588Pimbblet K. A., Drinkwater M. J., Hawkrigg M. C., 2004, MN-RAS, 354, L61Pimbblet K. A., Smail I., Edge A. C., O’Hely E., Couch W. J.,Zabludoff A. I., 2006, MNRAS, 366, 645Pimbblet K. A., Jensen P. C., 2012, MNRAS, 426, 1632Pinkney J., Roettiger K., Burns J. O., Bird C. M., 1996, ApJS,104, 1Poggianti B. M., Bridges T. J., Komiyama Y., Yagi M., CarterD., Mobasher B., Okamura S., Kashikawa N., 2004, ApJ, 601,197Poggianti B. M., et al., 2006, ApJ, 642, 188Popesso P., Biviano A., B¨ohringer H., Romaniello M., Voges W.,2005, A&A, 433, 431Popesso P., Biviano A., B¨ohringer H., Romaniello M., 2007a,A&A, 461, 397Popesso P., Biviano A., Romaniello M., B¨ohringer H., 2007b,A&A, 461, 411Quilis V., Moore B., Bower R., 2000, Sci, 288, 1617Reiprich T. H., B¨ohringer H., 2002, ApJ, 567, 716Rood H. J., Page T. L., Kintner E. C., King I. R., 1972, ApJ,175, 627Rusin D., et al., 2003, ApJ, 587, 143Sadat R., Blanchard A., Kneib J.-P., Mathez G., Madore B., Maz-zarella J. M., 2004, A&A, 424, 1097Shang C., Scharf C., 2009, ApJ, 690, 879Smith R. M., Driver S. P., Phillipps S., 1997, MNRAS, 287, 415Spergel D. N., et al., 2007, ApJS, 170, 377Stanford S. A., Eisenhardt P. R. M., Dickinson M., 1995, ApJ,450, 512Stott J. P., et al., 2012, MNRAS, 422, 2213Stott J. P., Pimbblet K. A., Edge A. C., Smith G. P., WardlowJ. L., 2009, MNRAS, 394, 2098Stott J. P., Smail I., Edge A. C., Ebeling H., Smith G. P., KneibJ.-P., Pimbblet K. A., 2007, ApJ, 661, 95Tempel E., Einasto J., Einasto M., Saar E., Tago E., 2009, A&A,495, 37 Tonnesen S., Bryan G. L., van Gorkom J. H., 2007, ApJ, 671,1434Urquhart S. A., Willis J. P., Hoekstra H., Pierre M., 2010, MN-RAS, 406, 368Valtchanov I., Murphy T., Pierre M., Hunstead R., L´emonon L.,2002, A&A, 392, 795van Dokkum P. G., Stanford S. A., Holden B. P., Eisenhardt P. R.,Dickinson M., Elston R., 2001, ApJ, 552, L101van Dokkum P. G., van der Marel R. P., 2007, ApJ, 655, 30Vikhlinin A., Forman W., Jones C., 1997, ApJ, 474, L7Vikhlinin A., et al., 2009, ApJ, 692, 1033Visvanathan N., Sandage A., 1977, ApJ, 216, 214White S. D. M., Briel U. G., Henry J. P., 1993, MNRAS, 261, L8Wilman D. J., Erwin P., 2012, ApJ, 746, 160Yamanoi H., et al., 2012, AJ, 144, 40Zabludoff, A. I., Huchra, J. P., Geller, M. J., 1990, ApJS, 74, 1Zel’Dovich Y. B., 1970, A&A, 5, 84c (cid:13)000