HST DETECTION OF SPIRAL STRUCTURE IN TWO COMA CLUSTER DWARF GALAXIES1 Alister W. Graham Department of Astronomy, University of Florida, P.O. Box 112055, Gainesville, FL 32611, USA [email protected] Helmut Jerjen Research School of Astronomy and Astrophysics, Australian National University, Private Bag, Weston Creek PO, ACT 2611, Canberra, Australia and Rafael Guzm´an Department of Astronomy, University of Florida, P.O. Box 112055, Gainesville, FL 32611, USA ABSTRACT We report the discovery of spiral-like structure in Hubble Space Telescope images of two dwarf galaxies (GMP 3292 and GMP 3629) belonging to the Coma cluster. GMP 3629 is the faintest such galaxy detected in a cluster environment, and it is the first such galaxy observed in the dense Coma cluster. The large bulge and the faintness of the broad spiral-like pattern in GMP 3629 suggests that its disk may have been largely depleted. We may therefore have found an example of the “missing link” in theories of galaxy evolution which have predicted that dwarf spiral galaxies, particularly in clusters, evolve into dwarf elliptical galaxies. Subject headings: galaxies: dwarf — galaxies: elliptical and lenticular, cD — galaxies: formation — galaxies: individual (GMP 3292, GMP 3629) — galaxies: spiral — galaxies: structure 1. Introduction happen when unsharp masking of a galaxy’s im- arXiv:astro-ph/0308241v1 14 Aug 2003 age reveals features indicative of a disk, such as Evidence for embedded, geometrically-flat, stel- bars and/or spiral patterns, or after a closer ex- lar disks has been found in a steadily increas- amination of a galaxy’s surface brightness profile ing number of objects that were once regarded reveals multiple component structure (e.g., Scorza as purely elliptical (E) galaxies (e.g., Capaccioli et al. 1998). It can, and has, also occurred after 1987, his section 5; Rix & White 1990; Vader & the inspection of kinematical data reveals signif- Vigroux 1991; Nieto et al. 1992; Cinzano & van icant rotation and/or bar-like dynamical behav- der Marel 1994; Jorgensen & Franx 1994; Sahu, ior (e.g., Carter 1987; Nieto, Capaccioli, & Held Pandey, & Kembhavi 1996). It has thus become 1988; Capaccioli & Longo 1994; Scorza & Bender a somewhat common trend for E galaxies to be 1995; Graham et al. 1998; Rix, Carollo, & Free- reclassified as lenticular (S0) galaxies. This can man 1999). 1Based on observations made with the NASA/ESA Hub- Similarly, the existence of previously unde- ble Space Telescope, obtained at the Space Telescope Sci- tected disks in dwarf elliptical (dE) galaxies are ence Institute, which is operated by the Association of Uni- now being realized. Although it has been known versities for Research in Astronomy, Inc., under NASA con- for nearly twenty years that some dE-like galaxies tract NAS 5-26555. 1 do have disk-like morphologies (dS0; Sandage & galaxy selection criteria is described in the follow- Binggeli 1984; Binggeli & Cameron 1991), what ing section, as is the image reduction process and is new — in addition to the rising number of analysis. Section 3 provides a brief quantitative disk detections — is that some dE-like galaxies analysis of the disks, and Section 4 discusses pos- actually have (stellar) spiral structures in their sible evolutionary scenarios for dwarf disk galaxies disk. After the initial surprise announcement of a in clusters. tightly-wound, two-armed spiral structure in the We take Coma to be at a distance of 100 Mpc −1 −1 ′′ dE galaxy IC 3328 (Jerjen, Kalnajs, & Binggeli and use H0=70 km s Mpc , 0.1 therefore cor- 2000b), Jerjen, Kalnajs & Binggeli (2001) and responds to 47 pc. Barazza, Binggeli, & Jerjen (2002) reported pre- viously undetected spiral structure and bars (i.e., 2. Galaxy Sample and Image Analysis disks) in four more Virgo cluster galaxies (2 dS0, 1 dE, and 1 low-luminosity E). De Rijcke et al. Galaxies meeting the following conditions — (2003) have additionally presented photometric discussed at length in Matcovi´c& Guzm´an (2003, and kinematic evidence for disks, and in one case in prep) — were selected from the Coma clus- spiral arms, in two edge-on Fornax cluster dS0 ter field catalog (Godwin, Metcalfe & Peach 1983; galaxies. The data to date suggests that up to hereafter GMP). All galaxies have positions within ′ ′ 20% of bright early-type dwarf galaxies in clusters the central 20 20 of the Coma cluster; 17.5 < × − may have disks. We report here on the first ever MB < 14.5; 0.2 < (U B) < 0.6 and 1.3 < − − detection of spiral-like structure in two dwarf, (B R) < 1.5; available HST WFPC2 images and − early-type galaxies residing in the densest clus- recessional velocities between 4,000 and 10,000 −1 ter environment studied so far, namely the Coma km s . The spectral analysis and recessional cluster. velocity derivation is also provided in Matcovi´c The presence, or at least detection, of (stellar) & Guzm´an (2003, in prep). The above require- spiral patterns in dwarf galaxies is a particularly ments were expected to result in the selection of rare phenomenon. Although dwarf versions of Sm Coma cluster dwarf elliptical galaxies, and we ob- and Irr galaxies have been known for a long time tained 18 such candidates. With the exception (e.g., van den Bergh 1960), referring to the early- of GMP 2960, there were no pre-existing mor- type Sa-Sc spiral galaxies, Ferguson & Sandage phological type classifications for these galaxies. (1991) wrote that “dwarf spiral galaxies do not ap- GMP 2960 (PGC 44707; Paturel et al. 1989) is pear to exist” (see also Sandage & Binggeli 1984 classified in NED as an S0 galaxy, and according and Sandage, Binggeli & Tammann 1985). They to the type-specific luminosity functions derived are a rare species; indeed, their very existence was from three clusters (Jerjen & Tammann 1997) we only recognized a few years ago (Schombert et al. conclude that GMP 2960 is either a low-luminosity 1995). Even then, Schombert et al. concluded that S0 or a bright dS0 galaxy. dwarf spiral galaxies only exist in the field. The The reduction process of the HST images is de- harsh environment within a galaxy cluster — due scribed in Graham & Guzm´an (2003). Briefly, to galaxy mergers, with each other or the intra- we used the iraf task crrej to combine the cluster medium, and/or strong gravitational tidal HST-pipelined exposures, which we then further interactions — is commonly thought to have led cleaned of cosmic rays using LACOS (L.A.COSMIC, to the destruction of the delicate spiral patterns van Dokkum 2001). Due to the stellar halos of in dwarf galaxies. A comparison of the number nearby galaxies, we used the wavelet decompo- of such objects in low- and high-density environ- sition method of Vikhlinin et al. (1998) to si- ments may shed light on the nature of their exis- multaneously subtract this non-uniform light and tence. the sky background. Foreground stars and over- By searching for signs of apparent spiral struc- lapping background galaxies were searched for, ture and/or bars in the optical images from the and masked out, before we performed any image sample of 18 dE galaxy candidates presented in analysis or surface brightness fitting. Graham & Guzm´an (2003), we explore here which In order to search for non-symmetric structures galaxies may have embedded stellar disks. The in the dwarf galaxy images, we subtracted the 2 axisymmetric component of the galaxy light (see large-scale disk displaying no obvious spiral struc- Jerjen et al. 2000b), leaving a “residual image”. ture. Neither the light-profile analysis in Graham Following Barazza et al. (2002) and De Rijcke & Guzm´an (2003) nor the present residual image et al. (2003), we have additionally used an un- analysis provide evidence to suggest that the re- sharp masking technique to verify the presence maining 14 galaxies are anything but nucleated of features such as bars or spiral arms, which dwarf elliptical galaxies. would indicate the presence of a flattened stellar disk. Although the majority of galaxies showed no 3. Quantitative Results sign of non-axisymmetric structure, two galaxies (GMP 3292 and GMP 3629) were found to pos- The inwardly extrapolated exponential disk in GMP 3292 has a central surface brightness sess flocculent spiral arms (Figures 1-2). Their −2 basic properties are given in Table 1. of µ0,F 606W =20.68 mag arcsec (Graham & Guzm´an, their table 2). Correcting this value From our previous analysis of the radial light- for Galactic extinction ( 0.02 mag; Schlegel, profiles (Graham & Guzm´an 2003), we had al- Finkbeiner, & Davis 1998),− (1 + z)4 redshift ready identified GMP 3292 as a likely bulge/disk dimming ( 0.10 mag), and K-correction (0.02 system due to a clear break in its surface bright- mag; Poggianti− 1997) gives a value of 20.58 mag ness profile marking the bulge/disk transition. arcsec−2. Assuming a B F606W color of 1.08 With regard to GMP 3629, we had remarked upon (Fukugita, Shimasaku, &− Ichikawa 1994) yields2 the possibility of an outer disk — not dominating −2 ′′ µ0 =21.66 mag arcsec , very close indeed to until radii greater than 10 (4.7 kpc) — but ,B the canonical Freeman (1970) value of 21.65 B- we could not and did not∼ confirm this due to the mag arcsec−2. The disk scale-length is 2.12′′, low surface brightness levels at these outer radii. which translates to 1 kpc. This is at the small We can however now confirm that both of these end of the range from∼ 1.0 to 2.5 kpc found in galaxies possess stellar disks as indicated by the Schombert et al.’s (1995) sample of dwarf spiral presence of a spiral pattern.