Properties of the Dwarf Galaxy Population in Galaxy Clusters
aa r X i v : . [ a s t r o - ph ] A p r Not to appear in Nonlearned J., 45.
Preprint typeset using L A TEX style emulateapj v. 08/22/09
PROPERTIES OF THE DWARF GALAXY POPULATION IN GALAXY CLUSTERS
R. S´anchez-Janssen , J. Alfonso L. Aguerri and C. Mu˜noz-Tu˜n´on Not to appear in Nonlearned J., 45.
ABSTRACTWe present the observational properties of the dwarf galaxy population ( M r > M ∗ r +1) correspondingto one of the largest samples of spectroscopically confirmed galaxy cluster members reported in theliterature. We have observed that red dwarf galaxies ( u − r > .
22) share the same cluster environmentas the brightest cluster members ( M r < − u − r < . u − r color of red dwarf galaxies was independent of their environment and similar to the color of redisolated dwarfs. Blue dwarf galaxies located outside r show similar u − r colors to those of the fieldpopulation, while strong reddening was observed toward the cluster center. We also present evidencethat the fraction of red to blue dwarf galaxies in clusters is larger in the innermost cluster regions.We conclude that the present red dwarf population observed in the central regions of nearby galaxyclusters could be related to the blue dwarf population observed in clusters at high redshift. Subject headings: galaxies: clusters: general — galaxies: evolution — galaxies: dwarf INTRODUCTION
The hierarchical galaxy formation scenario proposesthat massive galaxies grow continously through the ag-glomeration of smaller ones (White & Rees 1978). Inthis framework dwarf galaxies would then be the build-ing blocks of the galaxy formation process. Neverthe-less, their origin and evolution are still a matter of de-bate. In the standard scenario, dwarf galaxies are objectsformed via the gravitational collapse of primordial den-sity fluctuations. Once the first stars are formed, mech-anisms of energy feedback into the interstellar mediumare proposed in order to regulate the subsequent starformation or even to change the structure of the galaxy(Dekel & Silk 1986), a scenario supported by several ob-servational evidences (e.g. de Rijcke et al. 2005, and ref-erences therein). However, there are also several obser-vational proofs against the primordial origin of present-day dwarf galaxies in clusters. We can mention, forexample: the large scatter found in metallicities andages of dwarf galaxies in the Virgo and Fornax clus-ters (Conselice 2002); the different faint-end slope of thegalaxy luminosity function (LF) in clusters and in thefield (Blanton et al. 2005; Popesso et al. 2006); the vari-ation of the dwarf-to-giant ratio (DGR) as a functionof clustercentric distance observed in some nearby clus-ter (Phillipps et al. 1998); or the preferential location ofred bright and dwarf galaxies in overdense regions of theUniverse, in contrast to blue ones (Hogg et al. 2004).In high density environments such as galaxy clus-ters, there are many different physical mechanisms thatcan influence the evolution of cluster galaxies; e.g, ha-rassment (Moore et al. 1996), or ram pressure stripping(Quilis et al. 2000). In this scenario, dwarf galaxies canbe formed and destroyed, especially in the innermost re-gions of galaxy clusters, where strong tidal forces canproduce the destruction of galaxies, especially dwarfs Instituto de Astrof´ısica de Canarias. C/ V´ıa L´actea s/n, 38200La Laguna, Spain (Merritt 1984). In the Coma cluster, it has been ob-served that dwarf galaxies do not reach the very centralregions of the cluster, being distributed in a shell aroundthe cluster center (Trujillo et al. 2002). Dwarf galaxiescan also be created in galaxy clusters due to the evolutionof brighter ones. Strong tidal interactions between galax-ies and with the cluster potential can result in importantmass loss and transform bright galaxies into dwarfs (Mas-tropietro et al. 2005; Aguerri et al. 2004, 2005a).In this study we have measured the main observables ofthe dwarf galaxy population in one of the largest samplesof spectroscopically confirmed galaxy cluster members.The data will be presented in §
2; the results on the maindwarf galaxies characteristics are given in §
3, and in § THE DATA
A detailed description of the data, cluster global pa-rameters and cluster membership is given in Aguerriet al. (2007) (hereafter Paper I). We present here abrief summary. The data comprise all nearby clusterswith known redshift at z < . Only thoseisolated clusters with more than 30 galaxies with spec-troscopic data within the search radius and high spec-troscopic completeness ( > V c )and velocity dispersion ( σ c ) of the galaxies belonging to Throughout this study we have used the cosmological param-eters H = 75 km s − Mpc − , Ω m = 0 . Λ = 0 . S´anchez-Janssen et al.each cluster. Following the same approximation as Carl-berg et al. (1997), we also determined for each cluster theradius (r ) where the inner density is 200 ρ c , ρ c beingthe critical density of the Universe.The final catalog contains 89 isolated galaxy clusters.As mentioned before, the data were downloaded fromthe SDSS-DR4 database according to a metric criterion.This means that we are mapping different physical re-gions for each cluster. In order to avoid possible bi-ases, we have studied the r max /r ratio for each cluster,where r max is the maximum distance of a galaxy from itscluster center. We have found that all the clusters in oursample reach r max /r =2, and that 50% of them reachr max /r =5. The final sample of galaxies consist of 6880galaxies located within a radius 2 × r , and 10865 within5 × r . We builded an ensemble cluster by normalizingthe scales and velocities of each galaxy. Thus, the radialdistance of each galaxy to the cluster center was scaledby the r of the corresponding cluster, and the rela-tive velocity of each cluster galaxy was normalized bythe velocity dispersion of the cluster.All galaxy magnitudes were corrected for Galactic ab-sorption and k -corrected to the rest-frame at z = 0. Thegalaxies were classified according to their u − r color, asdescribed by Strateva et al. (2001). They showed thatthe distribution of SDSS galaxies in color–color diagramsis strongly bimodal, with an optimal color separation of u − r = 2 .
22 for early (red) and late-type (blue) galaxies. RESULTS
Fig. 1 shows the median clustercentric distance of thegalaxies as function of their absolute rest-frame r -bandmagnitude for objects residing within r < × r . The25% and 75% quartiles of the distributions are also over-plotted. Red galaxies always have smaller median clus-tercentric distances than blue ones. The median positonof galaxies brighter than M r ≈ − . r -band magnitude. It can be noticed that red galax-ies are always located in denser environments than blueones. Galaxies brighter than M r = − . r -band magnitude of the galaxies is also shownin Fig. 1. There is a segregation between red and blueobjects: red ones show lower velocity dispersions than We have considered the dwarf population as those galaxieswith absolute r − band magnitude fainter than M ∗ r +1, where M ∗ r − log ( h ) = − .
04, Blanton et al. (2005) blue ones. The brightest objects ( M r < − .
0) show asmaller velocity dispersion as they are brighter. Galax-ies with M r > − . Fig. 1.—
Median position (top), local projected density (middle)and velocity dispersion (bottom) of galaxy cluster members as afunction of their r -band absolute magnitude. In all panels, black,blue and red symbols represent the median values of these quanti-ties for the total, blue ( u − r < .
22) and red ( u − r > .
22) galaxypopulations, respectively. We have also overplotted the 25% and75% quartiles for the blue and red galaxy population.
We repeated the same calculations with galaxies lo-cated at r < × r in order to analyze the influence ofthe incompleteness in the mapping of the clusters, ob-taining the same trend in all relations. Another impor-tant incompleteness is the loss of faint galaxies towardshigher redshifts. Our galaxy cluster sample is completefor galaxies brighter than M r = − . z < . M r < − .
5. No changes in the trend of therelations were discovered.The galaxy density inside r in our clusters showslarge variations. This large cluster richness variationcould be responsible for the scatter of the measured quan-tities shown in Fig. 1. In order to account for this, thesample was split into two (high and low galaxy densityclusters) and Fig. 1 was recomputed. Nevertheless, bothsamples showed similar trends and scatter in the quan-tities reported in Fig. 1. This implies that the scattershown in Fig. 1 was intrinsic and not due to the ensem-warf galaxies in galaxy clusters 3ble of clusters with different richness.We have computed the DGR as a function of cluster-centric radial distance. The ratio was obtained consid-ering as dwarf galaxies as those with M r > M ∗ r + 1 andas bright ones those with M r < M ∗ r . The DGR has onlybeen computed using the previously mentioned completesubsample of clusters ( z < . r < (0 . − . r there are more red galaxies than blue ones. In contrast,blue galaxies are mostly located in the outermost regions( r > r ). Previous findings (Phillipps et al. 1998) haveshown that dwarf galaxies avoid high density environ-ments. We show for the first time the DGR from spec-troscopic observations and conclude that it is only theblue dwarf population that avoids cluster centers. Incontrast, the fraction of red dwarfs is constant (outside r/r > . u − r color of our dwarf population as a functionof the clustercentric radial distance. For comparison, wemeasured the mean rest-frame u − r color of a sampleof isolated dwarf galaxies (Allam et al. 2005). Figure 3shows the u − r color of blue and red dwarf galaxies as afunction of radius. Red dwarf galaxies in clusters have aconstant u − r color at all radii, and similar to the meancolor of isolated red dwarf galaxies. In contrast, the u − r color of blue dwarf galaxies depends on the position of thegalaxy in the cluster: galaxies located in the innermostregions of the cluster show redder u − r colors than thoselocated in the outermost regions. More precisely, bluedwarf galaxies located at r < r show redder colorsthan isolated blue dwarfs, while those located at r > r have similar u − r colors to isolated ones (see Fig. 3). DISCUSSION AND CONCLUSIONS
We grouped the galaxies in clusters into three differ-ent classes according to their absolute magnitude. Thefirst class was formed by the brightest cluster galaxies
Fig. 2.— (Top) dwarf-to-giant ratio, (middle) surface galaxy den-sity and (bottom) red-to-blue dwarf galaxy ratio as a function ofclustercentric distance. Blue and red colors as in Fig. 1.
Fig. 3.— u − r color as a function of clustercentric distanceof the dwarf galaxy cluster members (small points). Circles andtriangles represent the mean color corresponding to the red andblue dwarf galaxy populations, respectively. Their 25% and 75%quartiles are also overplotted. The shaded areas represent the 1 σ color distribution of red and blue dwarf field galaxies. ( M r < − . − . < M r < M ∗ + 1, and the third classrepresents the faintest galaxies in our sample, those with M r > M ∗ r + 1, which have been called dwarf galax-ies. The results presented in the previous sections showthat the brightest galaxies have a segregation with mag- S´anchez-Janssen et al.nitude in velocity and position: they show the lowestvelocity dispersion and are located closer to the clus-ter center in high density environments. As pointed outby (Biviano & Katgert 2004), these galaxies are not inequilibrium within the cluster potential. No other segre-gation with magnitude in the velocity space was foundfor the other two galaxy classes, except that blue galax-ies have a larger velocity dispersion than red ones. Thistrend has usually been interpreted in the past as a conse-quence of the different types of orbits presented by thesetwo classes of galaxies, with blue galaxies lying on moreradial orbits than red ones (Biviano & Katgert 2004).Dwarf galaxies also show a segregation with magnitudeand position. They are located closer to the cluster cen-ter as they are fainter, with red dwarf galaxies located athigher density environments than the blue ones. There-fore, the red dwarf population shares the same clusterenvironment as the brightest ones. Recent observationshave shown that galaxy clusters at z ≈ . ≈
10% of the light is lo-cated in the intracluster region (Arnaboldi et al. 2002;Aguerri et al. 2005b). Moreover, recent numerical simu-lations of ICL formation propose that this cluster com-ponent is mainly produced in major merger events thattake place during the formation of the brightest clustergalaxies. Mass stripped from galaxies only contibutes asmall fraction to the ICL (Murante et al. 2007). Oneprediction of the harassment model is that dwarf galax-ies should lie on low surface brightness tidal streams orarcs. However, Davies et al (2007) have found no evi-dence for any such features in a sample of dwarf galaxiesin the Virgo cluster.The most plausible explanation for the origin of theobserved red dwarf population located in the innermostregions of nearby clusters is that they are the evolvedblue dwarf galaxies observed in clusters at z ≈ . ∼
50 Myr; Quilis etal. 2000). The loss of the gas reservoirs will produce anoverall reddening of these galaxies, forming the red dwarfpopulation that we observe in nearby clusters.The authors would like to thank B. Moore, G. Yepes,and V. Quilis for useful discussion. We acknowledge fi-nancial support by the Spanish Ministerio de Ciencia yTecnolog´ıa grants AYA2007-67965-C03-01
REFERENCESAbell, G.O., Corwin, H.G., Jr.,& Olowin, R.P. 1989, ApJS, 70, 1Aguerri, J. A. L.,S´anchez-Janssen, R. & Mu˜noz-Tu˜n´on, C., A&A,471, 17 (Paper I)Aguerri, J. A. L., Iglesias-P´aramo, J., V´ılchez, J. M.,Mu˜noz-Tu˜n´on, C., & S´anchez-Janssen, R. 2005a, AJ, 130, 475Aguerri, J. A. L., Gerhard, O. E., Arnaboldi, M., Napolitano,N. R., Castro-Rodriguez, N., & Freeman, K. C. 2005b, AJ, 129,2585Aguerri, J. A. L., Iglesias-Paramo, J., Vilchez, J. M., &Mu˜noz-Tu˜n´on, C. 2004, AJ, 127, 1344Allam, S. S., Tucker, D. L., Lee, B. C., & Smith, J. A. 2005, AJ,129, 2062Arnaboldi, M., et al. 2002, AJ, 123, 760Ashman, K. M., Bird, C. M., & Zepf, S. E. 1994, AJ, 108, 2348Binggeli, B., Tammann, G. A., & Sandage, A. 1987, AJ, 94, 251Biviano, A., & Katgert, P. 2004, A&A, 424, 779 Biviano, A., & Girardi, M. 2003, ApJ, 585, 205Blanton, M. R., Lupton, R. H., Schlegel, D. J., Strauss, M. A.,Brinkmann, J., Fukugita, M., & Loveday, J. 2005, ApJ, 631, 208B¨ohringer, H., et al. 2000, ApJS, 129, 435Carlberg, R. G., Yee, H. K. C., & Ellingson, E. 1997, ApJ, 478,462Conselice, C. J. 2002, ApJ, 573, L5Davies, K. I., Roberts, S., & Sabatini, S., 2007, MNRAS, in pressDekel, A., & Silk, J. 1986, ApJ, 303, 39De Lucia, G., et al. 2007, MNRAS, 374, 809de Rijcke, S., Michielsen, D., Dejonghe, H., Zeilinger, W. W., &Hau, G. K. T. 2005, A&A, 438, 491Goto, T. 2005, MNRAS, 359, 1415Halliday, C., et al. 2004, A&A, 427, 397Hogg, D. W., et al. 2004, ApJ, 601, L29Kodama, T., et al. 2004, MNRAS, 350, 1005 warf galaxies in galaxy clusters 5warf galaxies in galaxy clusters 5