A&A 577, A59 (2015) Astronomy DOI: 10.1051/0004-6361/201425574 & c ESO 2015 Astrophysics Isolated ellipticals and their globular cluster systems III. NGC 2271, NGC 2865, NGC 3962, NGC 4240, and IC 4889,, R. Salinas1,2,A.Alabi3,4, T. Richtler5, and R. R. Lane5 1 Finnish Centre for Astronomy with ESO (FINCA), University of Turku, Väisäläntie 20, 21500 Piikkiö, Finland e-mail: [email protected] 2 Department of Physics and Astronomy, Michigan State University, East Lansing, MI 48824, USA 3 Tuorla Observatory, University of Turku, Väisäläntie 20, 21500 Piikkiö, Finland 4 Centre for Astrophysics and Supercomputing, Swinburne University of Technology, VIC 3122 Hawthorn, Australia 5 Departamento de Astronomía, Universidad de Concepción, Concepción, Chile Received 22 December 2014 / Accepted 21 February 2015 ABSTRACT As tracers of star formation, galaxy assembly, and mass distribution, globular clusters have provided important clues to our under- standing of early-type galaxies. But their study has been mostly constrained to galaxy groups and clusters where early-type galaxies dominate, leaving the properties of the globular cluster systems (GCSs) of isolated ellipticals as a mostly uncharted territory. We present Gemini-South/GMOS gi observations of five isolated elliptical galaxies: NGC 3962, NGC 2865, IC 4889, NGC 2271, and NGC 4240. Photometry of their GCSs reveals clear color bimodality in three of them, but remains inconclusive for the other two. All the studied GCSs are rather poor with a mean specific frequency S N ∼ 1.5, independently of the parent galaxy luminosity. Considering information from previous work as well, it is clear that bimodality and especially the presence of a significant, even dominant, popula- tion of blue clusters occurs at even the most isolated systems, which casts doubts on a possible accreted origin of metal-poor clusters, as suggested by some models. Additionally, we discuss the possible existence of ultra-compact dwarfs around the isolated elliptical NGC 3962. Key words. galaxies: elliptical and lenticular, cD – galaxies: star clusters: general – galaxies: structure 1. Introduction overlapping, have been put forward over the years (e.g., Ashman & Zepf 1992; Forbes et al. 1997; Côté et al. 1998; Beasley et al. The globular cluster systems (GCSs) of old, massive, elliptical 2002; Rhode et al. 2005; Muratov & Gnedin 2010; Elmegreen galaxies present an almost ubiquitous optical color bimodality et al. 2012; Tonini 2013). (e.g., Kundu & Whitmore 2001; Larsen et al. 2001), which is Since elliptical galaxies are known to mostly inhabit thought to correspond to a metallicity bimodality (e.g., Strader high-density environments (e.g., Dressler 1980), most of the et al. 2007; Usher et al. 2012; Brodie et al. 2012), with caveats GCS studies, and hence the observational constraints to these coming from possible nonlinearities of the color-metallicity theories, have naturally been focused on galaxy clusters (e.g., transformations that are introduced by horizontal-branch stars Peng et al. 2006; Harris et al. 2006; Strader et al. 2006; Liu et al. (Richtler 2006, 2013; Yoon et al. 2006; Chies-Santos et al. 2012; 2011). Blakeslee et al. 2012). This universal bimodality has obviously been one of the In the currently reigning paradigm of galaxy formation, ac- main aspects that the theories of GCS formation need to ad- cretion and merging play the central roles (e.g., Cole et al. 1994; dress. Many scenarios, not necessarily exclusive and somewhat De Lucia et al. 2006), and are strongly supported observationally (e.g., Ibata et al. 1994; Tal et al. 2009; van Dokkum et al. 2010). Based on observations obtained at the Gemini Observatory, which Simulations predict an accretion history for elliptical galaxies is operated by the Association of Universities for Research in in low-density environments that is different from the history Astronomy, Inc., under a cooperative agreement with the NSF on of their high-density siblings (e.g., Niemi et al. 2010); but the behalf of the Gemini partnership: the National Science Foundation impact that a low-density environment may have on the proper- (United States), the Science and Technology Facilities Council (United ties of a GCS has seldom been investigated (e.g., Gebhardt & Kingdom), the National Research Council (Canada), CONICYT Kissler-Patig 1999), although it could give important evidence (Chile), the Australian Research Council (Australia), Ministério da to distinguish between accretion-driven (e.g. Côté et al. 1998; Ciência, Tecnologia e Inovação (Brazil) and Ministerio de Ciencia, Hilker et al. 1999) and in situ (Forbes et al. 1997) scenarios. Tecnología e Innovación Productiva (Argentina). Globular cluster photometry is available in electronic form at the Only a handful of isolated elliptical galaxies (IEs) have thor- CDS via anonymous ftp to cdsarc.u-strasbg.fr (130.79.128.5) oughly studied GCS (e.g., Spitler et al. 2008) because there are or via so very few of them in the local environment. No field ellipticals http://cdsarc.u-strasbg.fr/viz-bin/qcat?J/A+A/577/A59 are listed, for example, in the extensive compilation of Brodie Appendices are available in electronic form at & Strader (2006, see their Table 1), and only a handful can be http://www.aanda.org seen in the newer compilation of Harris et al. (2013). Recently, Article published by EDP Sciences A59, page 1 of 18 A&A 577, A59 (2015) Table 1. Basic data of the galaxies presented in this work. T − Name RA Dec lbType B0 (m M)0 MB fov (mag) (mag) (mag) (kpc) R. Salinas et al.: Isolated ellipticals and their globular cluster systems. III. Table 2. Observations log. science frames were retrieved from the Gemini Science Archive. These frames were reduced following the same procedure as for Name Obs. date Exp. time (s) FWHM () the science images, with the exception of the fringe correction, (dd.mm.yyyy) g i g i which was deemed to be unnecessary given the short exposures ∼ i NGC 2271 01.02.2012 9 × 419 9 × 315 0.73 0.74 ( 5s)inthe filter. The number of standard stars measured for NGC 2865 25.01.2012 6 × 480 6 × 340 0.63 0.50 each night are listed Table 3. The transformation equations to the 26.01.2012 6 × 490 6 × 350 standard system used were × × NGC 3962 22.02.2012 6 610 6 440 0.73 0.60 = + + − + − NGC 4240 02.02.2012 5 × 422 5 × 275 0.80 0.58 gobs gstd Zg Kg(X 1) bg(g i )std (1) IC 4889 17.04.2012 3 × 575 5 × 355 – 0.75 = + + − + − 01.05.2012 3 × 575 1 × 355 0.49 – iobs istd Zi Ki(X 1) bi(g i )std, (2) where X denotes the airmass of the observation. Coefficients Notes. The FWHM was measured on the combined images. were obtained using IRAF/PHOTCAL. The extinction coeffi- cients were taken from the Gemini webpages and not fit, given the small range of airmasses covered by the observations. In ad- frames were retrieved from the Gemini Science Archive web- dition, when fewer than 20 stars were available in a field, the site. The FWHM measured on point sources in the final im- color terms were fixed to the values obtained for NGC 4240 ages is listed Table 2. The image quality in the IC 4889 ex- for which the largest number of standard stars was available. posures showed considerable variations between the two dates, Coefficients used for each galaxy are listed in Table 3. for this reason, the combined g image takes data only from the May 2012 observations, while the combined i was constructed with images taken in the night in April 2012 only. 2.1.2. Photometry completeness To study the level of completeness as a function of mag- 2.1.1. Stellar photometry nitude, artificial stars experiments were conducted using DAOPHOT/ADDSTAR. In each of the co-added frames (i.e., To facilitate the source detection, images were first median fil- those where the galaxy light was not yet subtracted), we tered using a box of 10 aside. This process removes the parent added 200 point sources covering a wide magnitude range. galaxy light (everywhere but in the innermost arcseconds), leav- These new frames were then processed in the same way as ing the point sources unaltered. A median value of the sky, de- previously described: galaxy subtraction with median filtering, termined from blank portions in the unsubtracted images distant source detection with SEXTRACTOR, and psf photometry with from the central galaxy, was then added back to the images to DAOPHOT. The procedure was repeated 50 times for a total keep the correct photon statistics during the detection and pho- of 10 000 artificial stars per galaxy per filter. An example of the tometry stages. fraction of recovered stars as a function of magnitude is plotted Object identification was carried out using SEXTRACTOR in Fig. 1 (top panel). (Bertin & Arnouts 1996). The choice of the detection filter The fraction of recovered stars was fitted with the interpola- can have important consequences on the number of identified tion formula, sources; while a Gaussian filter provides a better detection rate ⎛ ⎞ for the faintest sources, a “Mexican hat” filter gives better results 1 ⎜ α(m − m ) ⎟ = ⎜ − 50 ⎟ for sources very close to saturated stars or to the center of the f ⎝1 ⎠ , (3) 2 1 + α2(m − m )2 galaxy-unsubtracted residual. Given the relatively low Galactic 50 latitude of some of the targets (NGC 2271 and NGC 2865) which gives a good analytical description of the complete- and the clumpy nature of some galaxy residuals (for example, ness, and which has been used for many GC studies (e.g., IC 4889, see Sect. 6), the Mexican hat filter is the most suitable McLaughlin et al.