The Next Generation Virgo Cluster Survey. IV. NGC 4216: A Bombarded Spiral in the Virgo Cluster
Sanjaya Paudel, Pierre-Alain Duc, Patrick Cote, Jean-Charles Cuillandre, Laura Ferrarese, Etienne Ferriere, Stephen D. J. Gwyn, J. Christopher Mihos, Bernd Vollmer, Michael L. Balogh, Ray G. Carlberg, Samuel Boissier, Alessandro Boselli, Patrick R. Durrell, Eric Emsellem, Lauren A. MacArthur, Simona Mei, Leo Michel-Dansac, Wim van Driel
aa r X i v : . [ a s t r o - ph . C O ] F e b D RAFT VERSION O CTOBER
3, 2018
Preprint typeset using L A TEX style emulateapj v. 5/2/11
THE NEXT GENERATION VIRGO CLUSTER SURVEY. IV. NGC 4216: A BOMBARDED SPIRAL IN THE VIRGOCLUSTER * S ANJAYA P AUDEL † , P IERRE -A LAIN D UC , P ATRICK C ÔTÉ , J EAN -C HARLES C UILLANDRE , L AURA F ERRARESE , E TIENNE F ERRIERE , S TEPHEN
D. J. G
WYN , J. C HRISTOPHER M IHOS , B ERND V OLLMER , M ICHAEL
L. B
ALOGH , R AY G. C
ARLBERG ,S AMUEL B OISSIER , A LESSANDRO B OSELLI , P ATRICK
R. D
URRELL , E RIC E MSELLEM , L
AUREN
A. M AC A RTHUR , S IMONA M EI , L EO M ICHEL -D ANSAC , W IM VAN D RIEL , Laboratoire AIM Paris-Saclay, CNRS/INSU, Université Paris Diderot, CEA/IRFU/SAp, 91191 Gif-sur-Yvette Cedex, France National Research Council of Canada, Victoria, BC, V9E 2E7, Canada Canada-France-Hawaii Telescope Corporation, Kamuela, HI 96743, USA Department of Astronomy, Case Western Reserve University, 10900 Euclid Avenue, Cleveland, OH 44106, USA Observatoire Astronomique, Université de Strasbourg & CNRS UMR 7550, 11 rue de l’Université, 67000 Strasbourg, France Department of Physics and Astronomy, University of Waterloo, Waterloo, Ontario, N2L 3G1, Canada Department of Astronomy and Astrophysics, University of Toronto, Toronto, ON M5S 3H4, Canada Aix Marseille Université, CNRS, LAM (Laboratoire d’Astrophysique de Marseille) UMR 7326, 13388 Marseille, France Department of Physics and Astronomy, Youngstown State University, One University Plaza, Youngstown, OH 44555, USA Université de Lyon 1, CRAL, Observatoire de Lyon, 9 av. Charles André, 69230 Saint-Genis Laval; CNRS, UMR 5574; ENS de Lyon, France European Southern Observatory, Karl-Schwarzchild-Str. 2, D-85748 Garching, Germany GEPI, Observatoire de Paris, CNRS, Université Paris Diderot, 5 Place J. Janssen, 92190 Meudon Cedex, France
Draft version October 3, 2018
ABSTRACTThe final stages of mass assembly of present-day massive galaxies are expected to occur through the accretionof multiple satellites. Cosmological simulations thus predict a high frequency of stellar streams resulting fromthis mass accretion around the massive galaxies in the Local Volume. Such tidal streams are difficult to observe,especially in dense cluster environments, where they are readily destroyed. We present an investigation intothe origins of a series of interlaced narrow filamentary stellar structures, loops and plumes in the vicinity of theVirgo Cluster, edge-on spiral galaxy, NGC 4216 that were previously identified by the Blackbird Telescope.Using the deeper, higher-resolution and precisely calibrated optical CFHT/MegaCam images obtained as part ofthe Next Generation Virgo Cluster Survey (NGVS), we confirm the previously identified features and identify afew additional structures. The NGVS data allowed us to make a physical study of these low-surface brightnessfeatures and investigate their origin. The likely progenitors of the structures were identified as either alreadycatalogued VCC dwarfs or newly discovered satellites caught in the act of being destroyed. They have thesame g - i color index and likely contain similar stellar populations. The alignment of three dwarfs along anapparently single stream is intriguing, and we cannot totally exclude that these are second-generation dwarfgalaxies being born inside the filament from the debris of an original dwarf. The observed complex structures,including in particular a stream apparently emanating from a satellite of a satellite, point to a high rate ofongoing dwarf destruction/accretion in the region of the Virgo Cluster where NGC 4216 is located. We discussthe age of the interactions and whether they occurred in a group that is just falling into the cluster and showssigns of so-called "pre-processing" before it gets affected by the cluster environment, or in a group whichalready ventured towards the central regions of Virgo Cluster. In any case, compared to the other spiral galaxiesin the Virgo Cluster, but also to those located in lower density environments, NGC 4216 seems to suffer anunusually high heavy bombardment. Further studies will be needed to determine whether, given the surfacebrightness limit of our survey, about 29 mag arcsec - , the number of observed streams around that galaxy isas predicted by cosmological simulations or conversely, whether the possible lack of similar structures in othergalaxies poses a challenge to the merger-based model of galaxy mass assembly. Subject headings: galaxies: clusters: individual (Virgo) - galaxies: dwarf - galaxies: evolution - galaxies:interactions INTRODUCTION
Dwarf galaxies play a key role in Λ CDM-large scalestructure formation scenarios, e.g. Dekel & Silk (1986);Navarro et al. (1996). Indeed, according to this model, thedominant process of galactic growth is the accretion of small * Based on observations obtained with MegaPrime/MegaCam, a jointproject of CFHT and CEA/DAPNIA, at the Canada-France-Hawaii Tele-scope (CFHT) which is operated by the National Research Council (NRC)of Canada, the Institut National des Sciences de l’Univers of the Centre Na-tional de la Recherche Scientifique of France and the University of Hawaii. † Email: [email protected] galaxies, and the large stellar halos found around massivegalaxies such as our own Milky Way have been assembledthanks to the infall (and merging with) of many small dwarfgalaxies (Bullock & Johnston 2005). Prior to their capture,dwarf galaxies are often disrupted through large tidal forcesgenerated by their host galaxies, which form stellar streamsand shells around their hosts (See review by Duc 2012, andreferences therein). Although such fine structures have a life-time of a few Gyr in isolated environments, they are muchshorter lived in dense cluster environments, where they in-teract with the cluster potential and are quickly dispersed Paudel et al.out (Mihos 2004; Tal et al. 2009; Adams et al. 2012). Dis-rupted satellites may also be contributing stars to the intra-cluster light, where the debris generated by numerous inter-actions over the lifetime of the cluster is observed as a vastdiffuse halo around the most massive galaxies (Rudick et al.2009; Mihos et al. 2005; Janowiecki et al. 2010). On the otherhand, the cores of tidally stripped galaxies of intermediatemass may become compact ellipticals (cEs), and the nucleiof tidally stripped dwarfs may become ultra-compact dwarfgalaxies (Sasaki et al. 2007; Forbes et al. 2003; Huxor et al.2011). The scenario is commonly known as tidal threshing(Bekki et al. 2003).Because of their being stretched out, disrupted dwarf galax-ies experience a strong dimming of their surface bright-ness and become very difficult to detect. Until recently,very few candidate “disrupted dwarfs" were known outsideof the Local Group. With the advent of sophisticated in-struments and the development of data analysis techniquesoptimized to detect low surface brightness (LSB) struc-tures, the number of detections is growing (Forbes et al.2003; Martínez-Delgado et al. 2009, 2010; Duc et al. 2011;Miskolczi et al. 2011; Koch et al. 2012). However, most ofthe observed features – filaments, warps, plumes and shells –are left-overs from rather old collisions, as their progenitorsare no longer visible or identifiable. The deep imaging sur-vey of nearby spirals carried out by Martínez-Delgado et al.(2010) with robotic facilities such as the 0.5m Blackbird tele-scope show a number of cases of dwarf galaxies currentlyundergoing tidal disruption. These include a pair of dwarfsatellites of the Virgo spiral galaxy NGC 4216 which are as-sociated with extended tidal tails, whose brightest parts, cat-aloged as independent dwarf galaxies in Virgo Cluster Cat-aloge (VCC; Binggeli et al. 1985), are visible on shallowerSDSS images (Miskolczi et al. 2011).In this paper we revisit this system, using images obtainedas part of the Next Generation Virgo Cluster Survey (NGVS,Ferrarese et al. 2012). The excellent image quality of the sur-vey allowed us to disclose new objects, especially a coupleof filaments and disrupted dwarfs, and to perform a surfacephotometry and color analysis.NGVS observations and data reduction are presented inSect. 2. The results, in particular the photometric propertiesof the newly discovered features, are detailed in Sect. 3. InSect. 4, we investigate the origin of the streams, their age andfate. From a global view provided by the NGVS, we considerthe overall uniqueness of the NGC 4216 system within theVirgo Cluster. OBSERVATIONS AND DATA REDUCTION
The observations were carried out as part of the NGVSdeep imaging survey using the MegaCam instrument(Boulade et al. 2003) on the 3.6m Canada-France-HawaiiTelescope (CFHT). The observational strategy, data reduc-tion procedure and data quality control are described inFerrarese et al. (2012). In brief, the multi-band NGVS surveybenefits from the very wide field of view of the camera ( ∼ ◦ × ◦ ), and its very good image quality (0.6"-0.9") obtainedthrough optimizing the observing conditions. The images areprocessed with a dedicated data reduction pipeline, Elixir-LSB, which is explicitly designed to optimize the detectionof very low surface brightness structures such as intra-clusterlight (ICL), faint stellar streams and ultra-faint dwarf galax-ies (Magnier & Cuillandre 2004; Cuillandre et al., in prepa-ration). The output products of Elixir-LSB are single frames TABLE 1B
ASIC PROPERTIES OF
NGC 4216
AND ITS COMPANION GALAXIES .Galaxy RA Dec M B V d Morph. o o mag kms - kpcNGC 4216 183.976 13.149 - - - - - - OTE . — The coordinates, heliocentric radial velocities, V , and mor-phologies are from NED. The absolute B -band magnitudes are derived fromthe B-band magnitudes listed in the Goldmine database a . An average dis-tance for the Virgo Cluster of 16.5 Mpc (corresponding to 80 pc arcsec - )was assumed, and d is the projected distance on the sky from the center ofNGC 4216. a http://goldmine.mib.infn.it and stacks in which the sky background has been modeledand removed (Gwyn 2008; Gwyn et al. in preparation). Fur-ther image analysis has been done using standard routines ofIRAF and IDL.For the study, we made use of NGVS images obtained inthe g ′ , i ′ and z ′ bands . Total exposure times were 0.88, 0.57and 1.2 hours respectively. Since the g -band image is the mostsensitive and the cleanest, i.e., it contains the least number ofartifacts such as CCD imprints and reflection halos (see be-low), it was preferentially used for our photometric analysis.Surface photometry of very low surface brightness objectsis notoriously difficult. The precision of their measured pa-rameters is affected by several factors, including the accuracyof the photometric calibration, sky background variation andartifacts due to imperfect optics. The photometric calibra-tion of the NGVS has been done using the SDSS photome-try with a level of accuracy corresponding to a 1% variationin the photometric zeropoint (see Ferrarese et al. 2012). Atthe surface brightness limit of the survey (29 mag arcsec - inthe g -band), the main limitations to the LSB structure analy-sis and photometry are due to the reflection halos around thebright sources (stars but also the nuclei of bright galaxies withsteeply rising surface brightness profiles), which show up aslarge disks that cannot easily be removed from the images(see an example of one of these halos in the upper-left cornerof Figure 1). We therefore exclude in our analysis the con-taminated areas where reflected light becomes dominant overthe actual LSB feature.Foreground stars and compact background galaxies wereremoved applying a ring filter to the images, and residualswere manually subtracted with the IRAF task imedit . Theapertures used to carry out the photometry of the faint fila-ments and plumes, shown in Figure 2, were defined visually astheir surface brightness is too low for an automatic detection.The errors in the photometric measurements are to a large ex-tent determined by those in the sky determination. Although amaster sky background is subtracted from each image, for ouraperture photometry measurements we sampled the sky back-ground at multiple positions around the targets. The numberof sky probes ranged between 5 and 100, depending on thesize of the objects. Each individual sky region consisted of20 ×
20 pixel boxes. The median value and errors were then Image Reduction & Analysis Facility Software distributed by NationalOptical Astronomy Observatories, which are operated by the Association ofUniversities for Research in Astronomy, Inc., under co-operative agreementwith the National Science Foundation noted as g , i , z in the rest of the paper estruction of dwarf galaxies around NGC 4216 3computed. RESULTS
NGC 4216 and its environment
Figure 1 shows a composite MegaCam image of the fieldaround NGC 4216 (cf. Fig. 1c in Martínez-Delgado et al.(2010)). The regions with the highest surface brightnessare displayed in “true color" (combination of g , i and z band images) while for the low surface brightness regions, amonochromatic g –band image is shown.NGC 4216 is a barred spiral galaxy – it is classified asSAB(s)b in the RC3 cataloge (de Vaucouleurs et al. 1991) –located in the outskirts of the Virgo Cluster at an angular dis-tance of ∼ . It hasan edge-on orientation with an inclination angle i = 85 ◦ . Asshown in Figure 1, its central regions are characterized by aprominent dust lane (Chung et al. 2009) and a red and com-pact bright nucleus. The galaxy is surrounded by a very dif-fuse stellar halo, and complex filamentary structures, includ-ing streams and plumes. The investigation of their nature isthe focus of this study.The H I map of NGC 4216 from the VLA (Chung et al.2009) indicates an H I extent which is somewhat smaller thanin the optical: the ratio of the H I and optical B-band radii,measured respectively at the 1 M ⊙ pc - H I surface densitylevel and the 25 mag arcsec - isophotal level in the B-band, is D isoHI D B = 0.76.The low radial velocity of NGC 4216 (131 kms - , which issignificantly offset relative to the ∼ - mean veloc-ity of the cluster itself) indicates that it is moving relativelyfast with respect to the cluster center. It is the most mas-sive galaxy within a circle of 400 kpc projected radius; thenearest galaxy of higher mass is M 99 at a projected distanceof 420 kpc. Within that field of view, there are 2 early-typegalaxies, and 5 late-type star-forming galaxies with radial ve-locity differences less than ±
300 kms - with respect to that ofNGC 4216. The basic properties of NGC 4216 and its clos-est companions are listed in Table 1. In the close vicinity ofNGC 4216, towards the North lies VCC 165, at a projecteddistance of 19.2 kpc. VCC 165 is classified by Binggeli et al.(1985) as an S0 galaxy. Another companion is located fur-ther (56.5 kpc) to the North East, at the same radial velocity asVCC 165: the spiral galaxy NGC 4222. On our deep images,the stellar halo of the galaxy NGC 4222 overlaps with the re-flection halo of a bright star. South-East of NGC 4216, at58.0 kpc distance from its centre, at a similar radial velocity,lies a nucleated early type dwarf galaxy VCC 200 (Côté et al.2006). All these companions are identified on Figure 2. The streams around NGC 4216
Figure 2 identifies the main streams and remnants of dwarfgalaxies in the vicinity of NGC 4216 and VCC 165. TheNGVS images clearly show the main long filament F1, whichwas first detected by Martínez-Delgado et al. (2010). We alsosee the fainter structures present on the deep optical imageof Martínez-Delgado et al. (2010) such as the bend in F1 to-wards its Easter tip, the long narrow filament to the North ofthe system (F2) and the plume-like filaments that stick outEast of NGC 4216 (F5 and F6). Besides these previously The virial radius of Virgo’s A subcluster, centred on M87, is 5.4 ◦ (McLaughlin 1999). known structures, new features are detected, such as a fila-ment apparently wrapping around VCC 165 (F4), a loop Eastof NGC 4216 (F3) and a possible bridge between NGC 4216and VCC 200 (F7).With respect to previous surveys, the gain in spatial res-olution provided by CFHT/ MegaCam allows us to identifythe dwarf galaxies presumably at the origin of these fila-ments. Finally, the photometric calibration precision of themulti-band NGVS allowed us to determine the photometricproperties and colors of these various features, which so farhad not been possible. The photometric data on the streamsand progenitors are presented respectively in Table 2 & 3.The most prominent tidal stream, F1, East of NGC 4216,has an angular extent of ∼ ∼
60 kpc at the assumed mean distance of the clus-ter, 16.5 Mpc (Mei et al. 2007). The filament starts from theSouth-East corner of NGC 4216 and near its easternmost tipit makes a sudden turn towards the West-North with an an-gle of about 120 degree. Although further away it gets lostwithin the halo of a bright red star, when it emerges again,it seems to join the filament F2, the second longest filamentin the field, North of NGC 4216. Unfortunately, the stellarhalo glare in the region between the two filaments does notallow us to say with certainty that a physical connection ex-ists. No such bridge is visible on slightly shallower image ofMartínez-Delgado et al. (2010), where the stellar halo is lessobtrusive. The width of the stellar stream F1 appears constantalong most of its length, about ∼ - .The NGVS image discloses another rather broad stream,F3, with a surface brightness below 28.5 mag arcsec - (SeeFigure 3), which sticks out of NGC 4216, then bends to crossfilament F1 and finally falls back towards the host galaxy, nearF7.The very broad filament F7 points in the direction of thedwarf companion VCC 200, but a physical connection be-tween the two galaxies cannot be firmly established. Theplume-like stellar structures which apparently stick out fromthe main body of NGC 4216 on its western side (F5 and F6)may in fact be filaments which connect to F1–F3 behind thegalaxy. They have a nearly uniform surface brightness of 27.5mag arcsec - .Other filamentary structures can be seen (best on Figure 6)around VCC 165, the closest companion of NGC 4216. Onthe displayed g-band image, a model of VCC 165 obtainedwith the IRAF task bmodel has been subtracted to furtherenhance the filaments. The most prominent one, the 6-kpclong structure labeled F4, wraps around the dwarf companionon one side, and points towards the galaxy center on theother. There is a further hint that the galaxy extends furthertowards the East in the direction of the spiral. VCC 165itself, despite its apparent proximity to NGC 4216, does notexhibit strong signs of morphological perturbation. The mapobtained subtracting the ellipse model (see Figure 6) does notshow prominent residuals in its outer regions. In the innerregions, spiral-like features (or narrow dust lanes) may beobserved.We measured the g - i colors for the most prominent fea-tures. The i –band photometric measurements were particu- Paudel et al. F IG . 1.— Composite NGVS image of the field around NGC 4216. A monochromatic g –band image for which the faintest point-like stars have been subtractedis shown with a grey scale. For regions above a surface brightness level ∼
24 mag arcsec - , and in empty sky regions far from the main galaxy, a true color image(composite of g , i , z -band images) is superimposed. North is up, East is left and the field of view is 20 ×
17 arcmin. The image in the inset on the upper rightcorner is a zoom towards VCC 165, the closest companion of NGC 4216. larly delicate as this band shows several residuals at low sur-face brightness levels, such as chip gaps. The contaminationby stellar halos is also more severe in this band. Even the nu-cleus of NGC 4216 generates an artificial, disk–like structurewith a radius of nearly 3 arcmin, which is almost invisible inthe g –band, but becomes noticeable in the i and z bands. Re-gions affected by the stellar or galactic reflection halos wereexcluded from our analysis. The colors are listed in Table 2.The g - i color for F1 is uniform along its length and has anaverage value g - i = 0.9 ± . Stellar cores within the filaments
Several objects of higher surface brightness are visible to-wards the streams: such cores are likely the remnants oftidally disrupted dwarf galaxies and as such appear as goodprogenitor candidates of the filaments’ stellar material.As shown in Figures 2 and 3, the most extended filament,F1, hosts three relatively compact stellar objects. The mostluminous object, A, is also the most extended (see Figure 5).It has a ∼ - central g -band surface bright-ness, which in principle is high enough to have it included inthe Virgo Cluster Catalog (VCC 197). Modeling the galaxywith ellipses (using the IRAF task ellipse and bmodel ), wefound that it is stretched out along filament F1 and containsa bar-like structure along the minor axis, and a compact nu-estruction of dwarf galaxies around NGC 4216 5 F IG . 2.— Identification of the principal companion galaxies, stellar streams, plumes and progenitor candidates around NGC 4216. The apertures used to carryout the photometry are delineated on the displayed g –band image. Note that for objects A, B and C, the boundaries correspond to a surface brightness of 27.3mag arcsec - , the average surface brightness of filament F1. G1 and G2 are two selected regions in the extended stellar halo of NGC 4216. The dashed line nearthe extended stellar halo represents a possible, but not yet proven, connection between F1 and F2. TABLE 2P
ROPERTIES OF THE FILAMENTS (F1-F7)
AND EXTERNAL HALO REGIONS (G1-G2) OF NGC 4216.
Feature g - i < µ g > M B log( M ∗ )mag mag.arcsec - mag M ⊙ F1 0.9 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± OTE . — The color g - i was determined from aperture photometric mea-surements in areas not contaminated by instrumental artifacts. < µ g > is theaverage surface brightness within the areas defined in Figure 2, excluding theprogenitors (i.e. A, B & C). The associated errors are computed from the stan-dard deviation of the individual sky background measures. The absolute bluemagnitude in the B band, M B , was derived from the extrapolated observedmagnitude, i.e., the average surface brightness times the area, converting theMegaCam magnitudes to the standard Johnson system using the calibrationof Jordi et al. (2006) and assuming a distance of 16.5 Mpc (Mei et al. 2007).The stellar masses, M ∗ , were estimated from the extrapolated absolute mag-nitudes assuming the mass to light ratios of Bell et al. (2003). cleus. Modeling the galaxy with a Sérsic model with GALFIT(Peng et al. 2002), we obtained a best fit with an exponentialprofile and a very large effective radius of ∼ g - i colors of both A and B are in very good agreement with thatof filament F1, while C is 0.2 mag bluer. Given their ratherred colors, none of the galaxies show evidence of containingyoung stars. A is not detected on a deep 2600 sec archivalGALEX/NUV image (Boselli et al. 2011), while B showsvery weak UV emission. At Arecibo, the ALFALFA H I sur-vey (Haynes et al. 2011) did not detect any atomic hydrogenemission towards the three passive dwarfs. The much deeperAGES fields did not cover the NGC 4216 area (Taylor et al.2012a,b).The brightest condensation along Filament F4, West ofVCC 165, turns out to be a chance superposition of a back- Paudel et al. F IG . 3.— g -band surface brightness map of the field around NGC 4216. The surface brightness scale in mag arcsec - is shown to the right and the field of viewis the same as for Figure 2. The white contour corresponds to the faintest level of 21-cm H I line emission in NGC 4216 detected with the VLA by Chung et al.(2009).F IG . 4.— Variation of the median g –band surface brightness along the length of filament F1, from dwarf galaxy C near NGC 4216 to its easternmost tip where itgets lost within the halo of star. Photometric measurements were obtained collapsing the filament along its width and median-combining the data. The positionsof the dwarf galaxies A, B and C (see Fig. 2) are indicated. ground low–mass AGN at a redshift of 0.043, according to itsSDSS spectrum . The second brightest object in that regionis an extended structure that lies in the continuation of F4. The AGN, SDSS J121551.26+131303.4, belongs to a rich large-scalefilament of galaxies. Its emission lines classify it as a Seyfert 2. With anabsolute g –band magnitude of - < > -19 mag,did not reveal other objects with tidal tails apparently associated to them. Given its much higher surface brightness than that of filament(see Figures 5 and 7), we argue here that it could be in fact themain-body remnant of an almost completely disrupted dwarf,which we have labeled on the figures as D. The fact that Dand F4 have similar colors is consistent with this hypothesis.A small compact central core, labeled as N, can be seen to-wards D; it has the same color and could be the nucleus ofthe latter. The B –band absolute magnitude of D – within theboundaries defined in Figure 6 – is -11.3 mag, i.e. in betweenestruction of dwarf galaxies around NGC 4216 7 A B C F IG . 5.— g –band images of the disrupted dwarfs A, B, C and D. For objectsA, B and C, located along filament F1, the images have been rotated suchthat the local orientation of the filament is parallel to the horizontal axis.D is a disrupted dwarf galaxy candidate located West of VCC165. On thedisplayed image, the latter galaxy has been subtracted using an ellipse model.All images are displayed with the same intensity and spatial scales. that of dwarfs A and B.The mass of the stars still bound to the dwarfs (A+B+C+D)amounts to less than one tenth of the total stellar mass of thefilaments (F1-7). Assuming that all the stellar material in fila-ment F1 was pulled out from galaxy A, then the galaxy mustoriginally have had an absolute magnitude of M B = - DISCUSSION
In the Carnegie Atlas of Galaxies (Sandage & Bedke 1994),the spiral galaxy NGC 4216 is described as “one of the mostfamous galaxies in the sky because it is often used in text-books to illustrate the bulge, disk, dust content, and spiralpattern of typical luminous galaxies near the middle of theSb classification sequence". The deep NGVS images we havepresented here show many additional features around this pro-totypical galaxy, in particular a complex network of interlacedplumes, streams and filaments. Although a detailed interpre-tation is not straightforward, the large number of low sur-face brightness stellar tails is consistent with predictions fromnumerical simulations about the mass assembly of galaxies(Johnston et al. 2001; Bullock & Johnston 2005). This raisesthe question why all massive galaxies do not show a similar
TABLE 3P
ROPERTIES OF THE DWARF GALAXIES IDENTIFIED ALONG THE FILAMENTS . Object M B g - i e log(M ∗ ) R e mag mag – M ⊙ kpcA(VCC 197) - ± - ± - ± - ± - ± - ± OTE . — The absolute blue magnitude and colors were determined fromphotometric measurements made within the apertures defined in Figure 2.The ellipticities e were derived using the IRAF ellipse task. The effectiveradius are derived with a 2D fitting of the light profile with a Sérsic modelmade with GALFIT (Peng et al. 2002). D is the disrupted dwarf galaxy nearVCC 165 (see Fig. 6), while N is its candidate nucleus.F IG . 6.— g –band image of the field around VCC 165. The main featureshave been labeled: the streams F2 and F4, the progenitor candidate of F4, thedisrupted dwarf D, with its possible nucleus, N, and the background AGN.See the inset in Figure 1 for another view in color on this complex structure.F IG . 7.— Surface brightness profile along filament F4. The black solidcurved line in Fig 6 delineates the region traced for the profile. Paudel et al.level of fine-structures. In the following, we propose scenar-ios accounting for the various features seen around NGC 4216and discuss what is so special (or not) about this galaxy.
Real structures or artifacts
Disentangling real features from artifacts is essential to thegoals of this paper. Although multiple internal reflections inCCDs and stellar halos also produce low surface brightnessstructures, these instrumental features usually have somewhatregular geometric shapes, like disks, long thin straight linesor grids, while the shapes of the cosmic structures of inter-est to us are significantly different in appearance: broad fila-ments, arcs and irregular plumes. In the case of NGC 4216, anumber of the faint structures we noted are (barely) visible onSDSS images (Miskolczi et al. 2011) and more convincinglydetected by Martínez-Delgado et al. (2010).As a reminder, the principal features observed aroundNGC 4216 and possibly associated with collisional debris arethe following:(a) two long filaments (F1 and F2), with frail evidence thatthey are physically connected, forming a possible loop in theNorth-West quadrant of NGC 4216. Three dwarf galaxies (A,B and C) are found projected onto, and we argue, associatedwith filament F1;(b) a less extended very low surface brightness arc-like struc-ture east of NGC4216 (F3). The dwarf satellite B is found atthe intersection of F3 and F1;(c) a relatively high surface brightness filament wrappingaround the dwarf galaxy companion VCC 165, likely ema-nating from a newly discovered disrupted dwarf (D);(d) several broader and short filaments coming out from mainbody of NGC 4216 (F5-7).
Origin of the filamentary structures
Undoubtedly, all the stellar filamentary structures observedin the field of NGC 4216 have been shaped by tidal forces.However the origin of their stars may be debated. Given thehigh frequency of fine structures around the spiral, one maywonder if they were all produced by a single event: either anold merger with a relatively high mass companion, or ongo-ing tidal interaction with dwarf galaxies. In these two cases,the dwarfs seen along filament F1 would not be the originalprogenitors of its stars, but they would instead be tidal dwarfgalaxies (TDGs) born in situ from the collisional debris. Infact, the three aligned dwarfs A, B and C show similaritieswith the three also aligned TDG candidates recently discov-ered along a prominent tidal tail associated with the ellipticalgalaxy NGC 5557, which was interpreted as an old merger(Duc et al. 2011). However the latter are gas-rich, contrary tothe NGC 4216 dwarfs.NGC 4216 has at least one relatively massive companion,the edge-on spiral NGC 4222, and two luminous dwarf satel-lites, VCC 165 and VCC 200 (see Table 1). The latter twogalaxies are likely not massive enough to tidally disrupt theirhost (though the filament F7 pointing towards VCC 200 isintriguing). An on-going collision between the two spiralswould generate tidal tails in the plane of their stellar disks(see review by Duc & Renaud 2013). In fact, all the new fea-tures discovered around the galaxy (the stellar protuberances,plumes and filaments) are located along its minor axis, i.e.,perpendicular to its main plane, and not towards the outermostregions of its stellar disk plane, where the isophotes remainregular (see in particular Figure 1). In H I , Chung et al. (2009) did not find any clear indications of interactions between thetwo galaxies either, though they noted mild distortions in thekinematics at the edge of the disk in both systems. A ma-jor merger in NGC 4216 is thus very unlikely. Furthermore,the outer regions along the major axis of the galaxy, G1 andG2 (see Figure 2), are slightly bluer in g - i by 0.2 mag thanthe regions along the intergalactic stellar streams. This likelymeans that the stellar streams and outer disk of the galaxy donot share the same stellar populations, and thus that the starsin the filaments did not originate from the spiral.Alternatively, the filamentary structures may result fromdwarf galaxies tidally stirred by NGC 4216. In that case, thetails are shaped by both tidal forces and the orbit of the satel-lite around the host galaxy (Varghese et al. 2011). Inside sev-eral of the filaments, stellar bodies with higher surface bright-ness are found that are good filament progenitor candidates.The most promising one is object A, which shares with fila-ment F1 its orientation and color. Scrutinizing Figure 1, onemay see that the filament has a S-shaped structure close tothe dwarf, due to the presence of two slightly misaligned tailson each side of the galaxy, possibly the leading and trailingtidal tails. This provides an additional evidence that A is theprogenitor.On the other hand, it is hard to determine whether objectB belongs to filament F1 or F3, or if it is just a chancesuperimposition. Its main body is not aligned with any of thefilaments while its g - i color agrees with that of filament F1(the S/N of F3 is too low to determine its color). If objectsB, as well as C, are indeed associated with F1, their presencealong a single filament remains mysterious. It is unlikely thattheir common progenitor broke in two or even three stellarcores. Alternatively, if A, B and C were all pre-existingobjects, and subsequently tidally disrupted by NGC 4216,one would a priori expect them to be originally on differentorbits and therefore to generate filaments with differentorientations. Note however that according to cosmologicalmodels, satellite accretion occurs along privileged directions.The dwarfs might thus belong to a common cosmologicalstream. A chance superimposition of B and C is obviously anhypothesis worth exploring. Having spectroscopic data, andin particular information on the radial velocity and chemicalabundances, would help to determine the nature of the dwarfs.Alternatively, B and C might be substructures born in situin the tidal tail of A. Such a formation mechanism woulddiffer from the one at the origin of TDGs, as it would notinvolve a major wet merger. It would rather resemble the oneresponsible for the formation of the stellar clumps that arecommonly found in the tidal streams of Milky-Way GlobularClusters (e.g., Mastrobuono-Battisti et al. 2012). Note thatin our own local group, the distribution of dwarf satellitesinside narrow disks, the so-called disks of satellites, (e.g.Pawlowski et al. 2012; Ibata et al. 2013), is also puzzlingand no satisfactory solution to account for them has yet beenfound.Finally, the origin of the filamentary structures that do notshow any obvious progenitor, such as F2 (unless it is con-nected to F1 and shares the same progenitor), and of the threeplumes-like filaments (F5-7) is even less constrained. Theprogenitors might have been totally disrupted or their rem-nants might already have been swallowed by the massive spi-ral. In fact, object D near VCC 165 is likely a case of a dwarfin the process of total disruption. Whether it is disrupted fol-lowing its interaction with VCC 165 (around which its tidalestruction of dwarf galaxies around NGC 4216 9tail, F4, wraps) or with the more massive but slightly moredistant galaxy NGC 4216, or both, is unclear. VCC 165 itselfdoes not seem to be affected by gravitational interactions. An unusual on-going bombardment in a cluster ofgalaxies
The discovery of numerous filamentary structures aroundNGC 4216, and tidally stretched stellar clumps within them,revealed an impressive number of in-falling satellites caughtit the act of disruption. When has such a bombardment startedand where in the cluster did it occur?
A sub-structure within the Virgo Cluster?
Could NGC 4216 belong to a sub-structure, like a groupthat has been accreted by the Virgo Cluster? Within the clus-ter, NGC 4216 is located in a moderately dense environment:the local projected number density of galaxies is nearly halfthat of the cluster core. The spiral is the most luminous galaxywithin this sub-structure. For each Virgo galaxy more lumi-nous than M B = -18, we estimated in the SDSS dataset thenumber of companions within a 100 kpc radius and radial ve-locities of ±
300 kms - . NGC 4216 turns out to have a signif-icantly higher number of faint companions (M B > -18, withspectra available), compared to its fellow galaxies. The prob-ability of an increased number of dwarfs being demolished isthus higher there. The statistics on the ratio of dwarfs to giantsis still to be extended to the many faint and ultra-faint dwarfsthat are being discovered by the NGVS survey. Low veloc-ity collisions able to generate multiple tidal tails are expectedin groups. As a matter of fact, examples of tidal interac-tion and possibly ram–pressure have been observed in severalin-falling groups of clusters, which are suffering a so–called“pre-processing" (e.g., Cortese et al. 2006), during which theproperties of their galaxies are changed within the group be-fore being affected by the cluster environment. Time scales for the satellite accretion/destruction
A time scale for the satellite accretion/destruction processmay be obtained from the survival time of the debris: whereasaround isolated systems, streams may remain visible for sev-eral Gyrs, within a cluster, the life expectancy of tidal de-bris is largely reduced due to dynamical heating and mix-ing (Rudick et al. 2009). They should evaporate on a timescale shorter than the cluster crossing time, typically 1 Gyr(Trentham & Tully 2002). This suggests that the bombard-ment has either started within the Virgo Cluster itself or ina pre-processed group that was accreted by the cluster onlyrecently.The gas content of the NGC 4216 system somehow dif-fers from that of the pre-processed groups so far reportedin the literature. The filamentary structures detected inclaimed pre-processed groups (e.g., Cortese et al. 2006) areusually gas–rich, host H II regions, and are thus probably star-forming. There is no evidence for intergalactic H I gas aroundNGC 4216, at least in the ALFALFA single-dish survey The tails are red, presumably made of old stars. The dwarfsatellites around NGC 4216 had their gas stripped long ago, Considering that some large-scale gas features of low col-umn density such as the Virgo H I plume discovered by the WSRT(Oosterloo & van Gorkom 2005) have been missed in single dish surveys,deeper imaging studies with interferometers which will not resolve out dif-fuse gas will help to ensure the absence of intergalactic gas. either due to long-lasting interactions with the massive galax-ies in the group, or because globally the group members hadtheir gas reservoir stripped before the accretion event. Asnoted in Sect. 3.1, the H I disk of NGC 4216 seems trun-cated, which is normally interpreted as a consequence ofram pressure stripping in the cluster environment: the outergaseous halo of a spiral galaxy is stripped away when itpasses through a high density medium such as the core of acluster (Vollmer et al. 2001; Boselli & Gavazzi 2006). Thismight indicate that the NGC 4216 group already suffered aram-pressure stripping episode, and therefore that is has al-ready been in the cluster for a long time. The ram pressureneeded to strip a spiral down to the observed outer H I ra-dius R HI (15 kpc) is ∼ Σ HI V rot / R = ρ ICM V gal , where Σ HI isthe H I surface density (3 M ⊙ / pc ), V rot the rotation veloc-ity (250 km s - ) and V gal the galaxy velocity in Virgo cluster(1500 km s - ). This yields an ICM density, ρ ICM of 3 × - cm - . According to Schindler et al. (1999), such a density oc-curs at a cluster radius of about 0.5 Mpc or 1.7 degrees. Atthe present epoch, NGC 4216 is located at projected distanceof 2.5 and 3.6 degrees respectively from the central clustergalaxies, M86 and M87. Thus, the value of its H I truncationradius suggests a stripping episode that occurred at least 1 Gyrago, given the galaxy velocity, and closer to the cluster core,which would imply that the galaxy has ventured through thatenvironment in the past . The bombardment of NGC 4216and the multiple dwarf destruction must then have occurredlater on.Alternatively, the H I truncation of NGC 4216 might bepossible without invoking the Virgo Intra Cluster Medium,if the group were sufficiently massive and X–ray luminous(Sengupta et al. 2007; Kern et al. 2008). There is however noevidence for extended X-ray emission at the location of thegalaxy to support the hypothesis of an in situ ram pressureevent (See Figure 7 in Chung et al. 2009). Finally one shouldnote that one of the main characteristics of the gas property ofNGC 4216, more striking than the disk truncation, is its over-all low H I surface density for a spiral galaxy of its size. Sucha global depletion of H I might in fact be due to the effect ofturbulent viscous stripping (Nulsen 1982; Chung et al. 2009)rather than ram pressure stripping.Taking furthermore into account the fact that the companiongalaxy, NGC 4222, has a regular H I distribution, it is stilldifficult to firmly conclude that the group of NGC 4216 didexperience a ram pressure event during a past journey throughor near the Cluster Core. How unique is NGC 4216?
How special is NGC 4216 within the Virgo Cluster? Aspart of the NGVS project, we have identified in a systematicway low-surface brightness stellar tails, and cases of tidallydisrupted dwarfs. Our preliminary analysis over most of theVirgo Cluster area indicates that there are only a rather smallnumber of them, especially around spiral galaxies (Duc et al.,2013 in prep.). This may be an additional argument in agree-ment with a recent accretion of NGC 4216 as part of a group. The true distance and location of NGC 4216 within the Virgo Clus-ter is not known. An estimate of the distance based on Surface BrightnessFluctuations is only available for one possible group member, VCC 200.Blakeslee et al. (2009) determined a distance of 18.2 ± ∼ - – then the number of filaments andsatellites is expected to strongly increase according to numer-ical simulations (Bullock & Johnston 2005, Michel-Dansac etal., 2013, in prep.). CONCLUSIONS
We have presented a study of the properties and origins of acomplex series of interlaced narrow filamentary stellar struc-tures, loops and plumes located in the vicinity of the edge-onVirgo Cluster spiral galaxy NGC 4216. They were detectedon multi-band extremely deep optical Next Virgo ClusterSurvey (NGVS) images. Some streams had already beenidentified in previous surveys, e.g. Martínez-Delgado et al.(2010). Several others, with g –band surface brightness levelsfainter than 28.5 mag arcsec - are reported here for the firsttime.Whereas the tidal origin of the filamentary structures seemsrather obvious, determining the progenitors of their stars isless straightforward. The high resolution of the MegaCamcamera, besides its high sensitivity to low surface brightnessfeatures, allowed us to identify progenitor candidates and tofurther explore the complexity of the system. This is illus-trated by the detection of satellites of satellites: a filament wasfound wrapping itself around the luminous dwarf VCC 165,which itself is probably orbiting around NGC 4216. Thealignment of three dwarfs along an apparently single streamis also intriguing. Projection effects might be an explanationunless some of them are second-generation objects formed in-situ in tidal debris, or if all the dwarfs were accreted from thesame cosmological filament.The dominant spiral itself does not show any strongevidence of having experienced a recent merger or beingcurrently involved in a tidal interaction with a massivecompanion. The stellar streams most likely result fromthe disruption of dwarf satellites, some of which are stillvisible but look highly highly disrupted on the NGVS images.While dwarf galaxy accretion, with an associated high rate of satellite destruction and the formation of stellar streams, isa priori expected for massive galaxies – cosmological semi-analytical models and simulations predict that they growat low redshift through minor mergers (Naab et al. 2009;Khochfar et al. 2011; McLure et al. 2013) - their presencein a cluster environment is rather surprising. A possibleexplanation is that NGC 4216 is the central galaxy of a groupthat has recently fallen into the Virgo Cluster, and that themerging activity seen in its surroundings might be consideredas evidence for pre-processing. This scenario does not fullyexplain the apparently truncated H I distribution of the spiralgalaxy, however, if caused by a ram-pressure stripping event.The level of ram-pressure required to account for the H I deficiency might imply that the spiral has already passedcloser by the central regions of the cluster at least ∼ I disk. Follow-up observations, in particulara spectroscopic survey to get dynamical information on thesystem, would be needed to determine how long ago thegroup of NGC 4216, including its numerous satellites, joinedthe Virgo Cluster.We have carried out a preliminary comparison of the finestructure index of NGC 4216 with that determined in otherspiral galaxies in the Virgo Cluster, but also around massivegalaxies in the field and in groups. NGC 4216 seems to have aparticularly high number of streams typical of minor mergers,and associated dwarfs caught in the act of being disrupted.We intend to confirm this as part of other on-going deep sur-veys with the Megacam camera on the CFHT. The signifi-cance and implications of our results should also be probedusing numerical simulations of galaxy evolution which maytell whether NGC 4216 has indeed had an unusually rich re-cent accretion history, if on the contrary its fellow-membershave had a much less important minor merger activity thanpredicted, or whether our knowledge of stellar streams and itsassociated time scales should be revisited. ACKNOWLEDGMENTS
We wish to express our gratitude to the CFHT personnel forits dedication and tremendous help with the MegaCam obser-vations. Many thanks to the anonymous referee for his usefuland relevant comments. This work is supported in part bythe French Agence Nationale de la Recherche (ANR) GrantProgramme Blanc VIRAGE (ANR10-BLANC-0506-01), andby the Canadian Advanced Network for Astronomical Re-search (CANFAR) which has been made possible by fund-ing from CANARIE under the Network-Enabled Platformsprogram. J.C.M. thanks the NSF for support through grantAST-1108964. P.R.D acknowledges support from NSF grantAST-0908377. This research used the facilities of the Cana-dian Astronomy Data Centre operated by the National Re-search Council of Canada with the support of the CanadianSpace Agency. The authors further acknowledge use of theNASA/IPAC Extragalactic Database (NED), which is oper-ated by the Jet Propulsion Laboratory, California Institute ofTechnology, under contract with the National Aeronautics andSpace Administration.
REFERENCESAdams S. M., Zaritsky D., Sand D. J., Graham M. L., Bildfell C., HoekstraH., Pritchet C., 2012, AJ, 144, 128 Bekki K., Couch W. J., Drinkwater M. J., Shioya Y., 2003, MNRAS, 344,399 estruction of dwarf galaxies around NGC 4216 11estruction of dwarf galaxies around NGC 4216 11