DISCOVERY OF A POSSIBLY SINGLE BLUE SUPERGIANT STAR IN THE INTRA-CLUSTER REGION OF VIRGO CLUSTER OF GALAXIES

As galaxies fall into the huge gravitational potential of clusters of galaxies with high velocity, interaction with the hot intra-cluster medium strips off cool gas from the main body of the infalling galaxies into the intra-cluster region (Gunn & Gott 1972; Vollmer et al. 2001). Studying the fate of this stripped gas, ionization in contact with a million degree intra-cluster medium, or condensation to form new stars, is of great interest. The Virgo cluster is the closest massive cluster of galaxies (16.5 Mpc; Mei et al. 2007) and is an ideal laboratory to investigate details of such processes. We focus on IC 3418, a galaxy likely falling into the Virgo cluster for the first time at a very high speed (nearly 1000 km s⁻¹; Vollmer et al. 2001), especially on its 17 kpc long trail which is considered to have formed behind the galaxy due to ram pressure stripping (Hester et al. 2010; Fumagalli et al. 2011). The trail is manifested as a chain of blobs seen in both GALEX (Martin et al. 2005) UV continuum and Hα emission-line images (Hester et al. 2010; Fumagalli et al. 2011). Both of these emissions are clear signs of recent star formation, and therefore the trail is the best target to explore details of a young massive stellar population in the intra-cluster region. Stellar spectroscopy is unique and useful not only because it provides basic stellar properties (e.g., age, metallicity, luminosity), but also because we can explore a potentially new mode of star formation in this special condition. We expect that such star formation is quite different from those in the Milky Way, and such a study will eventually contribute to the understanding of galaxy evolution under cluster environment (Gunn & Gott 1972) and intra-cluster stellar population.

In this Letter, we report on our spectroscopic studies of SDSS J122952.66+112227.8 (hereafter SDSS J1229+1122), a compact source of optical emission within the star-forming trail of IC 3418.

The FOCAS spectrograph at the Subaru telescope (Kashikawa et al. 2002) was used on the night of 2011 April 27 in multi-slit spectroscopy mode to study the nature of the trail of IC 3418. One of our slitlets was put on a faint ($g^{\prime }_{\rm AB}=23.0$ mag) blue compact object catalogued earlier as SDSS J1229+1122 and located near the far end of the trail (Figures 1 and 2). Low-resolution spectrum was taken with an 08 wide slit and a "300B" grism for spectral resolution R(≡ λ/δλ) of ≃ 700 to cover 4000–7000 Å, and the total exposure time was 6000 s. Galactic extinction for E(B − V) = 0.03 mag was corrected based on the NASA Extragalactic Database (NED) information. The absolute flux scale of the spectrum was re-calibrated based on our photometric data (see below), and a small slit-loss correction (by 7%) was applied. Additionally, a shorter exposure (3600 s) but with relatively higher ("medium") spectral resolution (R ≃ 3750) spectrum was obtained with the same slit width and "VPH680" grism, resulting in a FWHM velocity resolution of ≃ 80 km s⁻¹. Low-resolution spectrum is shown in Figure 3, and measured spectroscopic properties on both spectra are summarized in Table 1.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Table 1. Photometric and Spectroscopic Characteristics of SDSS J1229+1122

Megacam Astrometry	R.A. (2000)	Decl. (2000)
	12h29m5269	+11°22'280
Megacam photometrya	Observed	Correctedb
u*	23.29 ± 0.02	23.15 ± 0.02
g'	22.96 ± 0.01	22.85 ± 0.01
i'	23.09 ± 0.02	23.03 ± 0.02
z'	23.16 ± 0.04	23.12 ± 0.04
Bessel photometryb,c	Apparent	Absoluted
V	22.85 ± 0.01	MV = −8.25
FOCAS low-resolution spectroscopy
Hα fluxb	(5.04 ± 0.08) × 10−17 erg s−1 cm−2
Hα equivalent width	−33.7 ± 0.7 Å
Hβ fluxb	(2.08 ± 0.22) × 10−17 erg s−1 cm−2
Hβ equivalent width	−6.7 ± 0.7 Å
Emission-line ratios	3σ upper limit
$\rm [{\rm O\,{\scriptsize{III}}}]\lambda$5007/Hβ	<0.28
$\rm [{\rm N\,{\scriptsize{II}}}]\lambda$6583/Hα	<0.028
$\rm [S\, \scriptsize{II}]\lambda$6717+6731/Hα	<0.031
FOCAS medium-resolution spectroscopy
Hα heliocentric velocity	−99 ± 6 km s−1
Hα intrinsic line width	164 ± 22 km s−1 FWHM

Notes. ^aIn Megacam SDSS (AB) system. ^bGalactic extinction corrected for E(B − V) = 0.03. ^cBessel system in Vega unit, converted based on our spectrum. ^dDistance modulus of m − M = 31.1 is used.

Download table as:  ASCIITypeset image

We also performed photometric analysis on archival images to supplement our spectroscopy (Figures 1 and 2). Canada–France–Hawaii Telescope (CFHT) Megacam (Boulade et al. 2003) images were taken from Megapipe (Gywn 2008) at the g', i', and z' bands, and another image at the u* band was locally stacked based on individually pre-calibrated archival images available at the CFHT Science Data Archive. All those images were originally taken for the New Generation Virgo cluster Survey (NGVS) project (Ferrarese et al. 2012). SDSS J1229+1122 has a faint nearby source within 1'' from our spectroscopic target source, which was not resolved in the Sloan Digital Sky Survey (SDSS) image. SDSS J1229+1122 was unresolved in both the i'- and z'-band images having the best FWHM seeing size of 07 (≃ 60 pc at the distance of the Virgo cluster). Multi-color photometry was performed through double-Gaussian fitting on the images to properly remove any contamination from the nearby source, and the same amount of Galactic extinction is corrected as for the spectra. In addition, an Hα image was obtained with EFOSC2 attached to the New Technology Telescope on 2011 April 29. A narrow "Ha#692" filter covering Hα within a velocity range of −780 to 2060 km s⁻¹ and a broadband "R#642" filter are used for stellar continuum-subtracted Hα line images. Total exposure times were 1800 s and 600 s for the Hα and R-band images, respectively. Flux calibration was performed with a spectroscopic standard star LTT3218. Our point source detection limit (3σ) is 1.8 × 10⁻¹⁶ erg s⁻¹ cm⁻². Note that this image is independent from that presented in Hester et al. (2010) and Fumagalli et al. (2011). Also, a deeper GALEX NUV image (1750–2800 Å) was created by combining the GALEX Ultraviolet Virgo Cluster Survey (Boselli et al. 2011) and the Deep Imaging Survey (Martin et al. 2005). The total integration time is 20672 s. All the images are shown in Figures 1 and 2, and optical photometry data (Table 1) are overlaid on the spectra (Figures 3 and 4).

Zoom In Zoom Out Reset image size

The low-resolution spectrum of SDSS J1229+1122 can be characterized by strong Hα (at equivalent width (EW) of −34 Å) and weak Hβ emission lines on blue continuum. The Hα line is slightly resolved with our medium-resolution spectrum, having a FWHM of ∼164 km⁻¹ (corrected for the instrumental resolution), but further line profile analysis was not possible. Our spectroscopic Hα flux (5.0 × 10⁻¹⁷ erg s⁻¹ cm⁻²) is below the detection limit of our Hα image (Figure 2). The Hα flux is ≲ 10 times fainter than nearby compact Hα emitters within the trail (Hester et al. 2010; Fumagalli et al. 2011; see also Figure 2). Our medium-resolution spectrum shows that the heliocentric velocity is −99 km s⁻¹, which is very close to that of IC 3418 (38 km s⁻¹; Gavazzi et al. 2004). Their similar recession velocities as well as spatial association of SDSS J1229+1122 with the trail of IC 3418 (Figure 1) strongly indicate that they are physically associated to each other within the Virgo cluster. At a distance of 16.5 Mpc (or m − M = 31.1; Mei et al. 2007), the absolute magnitude (M_V) is ∼ − 8.3 and the Hα luminosity is 1.6 × 10³⁶ erg s⁻¹. The Hα luminosity is about 5–100 times fainter than in known intra-cluster H ii regions around NGC 4388 (Gerhard et al. 2002; Cortese et al. 2004) and about 10 times fainter than those around the VCC 1249/M 49 merging system (Arrigoni Battaia et al. 2012) in the Virgo cluster. In one of these known H ii regions, where detailed spectroscopic analyses were made, only a few O-type stars were estimated to be powerful enough for the observed luminosity (Gerhard et al. 2002).

We find that the emission-line ratios of SDSS J1229+1122 are unusual for photo- or shock-ionized nebulae. Other than Hα and Hβ, no nebular forbidden lines were detected, including ones that are typically bright in those nebulae ([O iii] λ5007 Å, [N ii] λλ6548,6583 Å, and [S ii] λλ6717,6731 Å; Figure 3). Their upper limits were measured assuming the same line widths and redshift as Balmer emission lines (Table 1). Stringent upper limits on low ionization [N ii] and [S ii] and undetected higher ionization [O iii] lines strongly indicate that the emission originates from neither H ii regions nor shock-ionized nebulae (Veilleux & Osterbrock 1987) but from hot stellar atmosphere. We also note that other emission lines seen typically in luminous blue variable (LBV) stars besides Balmer lines, such as [He i] λ5878 Å and [Fe II] λ5018 Å, were not detected.

Such a luminous compact object with blue continuum should either be a young stellar association or a single blue (super)giant star. We first searched a stellar spectrum library (Pickles 1998) for matching the continuum (below Hα and excluding Hβ emission lines) and the broadband photometric data, and found that A2 supergiant (I) is the best match to the observation (Figure 4). A similar blue continuum can be reproduced by the A5 dwarf (V), but its u* flux is expected to be fainter than the observation and its intrinsically faint luminosity does not agree with the observation (Figure 4). Then we searched starburst99 models of instantaneous star formation (Leitherer et al. 1999) for the same purpose. Continuous star formation models are not considered since they should always show H ii-region-like spectrum, unlike our observation. We chose a model of 0.4 solar metallicity (Z = 0.008) based on metallically measurements with other intra-cluster H ii regions in the Virgo cluster (Gerhard et al. 2002; Arrigoni Battaia et al. 2012). The shape of the initial mass function (IMF) cannot be examined, and we only describe here cases with IMF slope (α) = 2.35 and M_upper = 100 M_☉ since the basic results hold with other IMF parameter sets. Although the model grid provided by Leitherer et al. (1999) is not fine enough to find statistically satisfactory models for all photometry data, we found three possible burst ages to match the observation: 6 Myr, 16 Myr, and 20 Myr (Figure 4). We found three solutions because, in addition to main-sequence stars that become monotonically redder with age, enhanced red supergiant contribution quickly makes the color redder around ∼10 Myr (or between 6 and 16 Myr in the assumed model parameters) when most O-type stars evolve away from the main sequence. Note that the very similar trend of color change is seen for all different IMF parameter sets. In summary, both possibilities of a blue supergiant (hereafter BSG) star and young (6–20 Myr) stellar association are equally good to reproduce the observed blue continuum spectrum and broadband photometry.

4.1. SDSS J1229+1122: A Compact Young Stellar Association or a Single Blue Supergiant Star?

4.2. Star Formation Characteristics within the Cluster Environment

Various kinds of intra-cluster stellar populations have been known in the Virgo cluster: in addition to diffuse intra-cluster optical continuum light (Mihos et al. 2005), individual sources like intra-cluster H ii regions (Gerhard et al. 2002; Cortese et al. 2004; Arrigoni Battaia et al. 2012), old (⩾10 Gyr) intra-cluster stars (Williams et al. 2007), and intra-cluster planetary nebulae (Feldmeier & Ciadullo 2004; Aguerri et al. 2005) are known. Our study, independent of the interpretation that SDSS J1229+1122 contains a single or a few more stars, has concluded that BSG reside in the intra-cluster region. Previously, BSG candidates have been photometrically identified in M 100 (Hill et al. 1998) in the Virgo cluster, and spectroscopically identified in NGC 3621 at a distance of 6.7 Mpc (Bresolin et al. 2001). Located in the Virgo cluster, and if confirmed by future observations to be a single star, SDSS J1229+1122 will be the most distant star discovered from spectroscopic observations. We demonstrated for the first time that stellar spectroscopy up to the distance of the Virgo cluster is indeed feasible as anticipated nearly two decades ago (Kudritzki et al. 1995; McCarthy et al. 1997; Bresolin et al. 2001; Kudritzki 2010). However, to reveal its true nature by detailed quantitative spectroscopic analyses, spectra with higher signal-to-noise ratios and higher resolution below Hβ (⩽4800 Å) as well as Hα are required. To exploit its full potential for accurate distance measurements and star formation studies in the Virgo cluster and beyond, we require next-generation giant telescopes such as Thirty Meter Telescope (TMT) and/or European Extremely Large Telescope (E-ELT; Kudritzki 2010).

Based in part on data collected at Subaru Telescope, which is operated by the National Astronomical Observatory of Japan. Based on observations obtained with MegaPrime/MegaCam, a joint project of CFHT and CEA/DAPNIA, at the Canada–France–Hawaii Telescope (CFHT), which is operated by the National Research Council (NRC) of Canada, the Institut National des Science de l'Univers of the Centre National de la Recherche Scientifique (CNRS) of France, and the University of Hawaii. This research used the facilities of the Canadian Astronomy Data Centre operated by the National Research Council of Canada with the support of the Canadian Space Agency. Funding for the SDSS and SDSS-II has been provided by the Alfred P. Sloan Foundation, the Participating Institutions, the National Science Foundation, the US Department of Energy, the National Aeronautics and Space Administration, the Japanese Monbukagakusho, the Max Planck Society, and the Higher Education Funding Council for England. Based on data obtained from the ESO Science Archive Facility under request number "Youichi Ohyama #35,191." Some of the data presented in this Letter were obtained from the Mikulski Archive for Space Telescopes (MAST). This research has made use of the NASA/IPAC Extragalactic Database (NED), which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration. This work is supported by grant NSC 100-2112-M-001-001-MY3 (Y.O.). A.H. is thankful to Inter-University Centre for Astronomy and Astrophysics (IUCAA), India, for a visit with financial support and to National Centre for Radio Astrophysics of Tata Institute of Fundamental Research (NCRA-TIFR), India, for a Visiting Astronomer position during which parts of the work were done.