MCG+08-22-082: A DOUBLE CORE AND BOXY APPEARANCE DWARF LENTICULAR GALAXY SUSPECTED TO BE A MERGER REMNANT

To perform detailed surface photometry on U141, we extensively use imaging from the SDSS Data Release 7 (DR7, Abazajian et al. 2009). Reduced and calibrated CCD images were acquired from the SDSS data archive server. These were observed in five optical bands, i.e., u, g, r, i, and z, with a total exposure time of 54 s in each band. The pixel scale is 0396 pixel⁻¹ and the average seeing in this field of view is 12 in the r-band, as we have measured from the full width at half maximum (FWHM) of randomly selected foreground stars. Details of the post-processing of the SDSS images are given in Pak et al. (2014). Since sky background subtraction in the SDSS pipeline processed images is not reliable (Lisker et al. 2006b), we optimized the post-processing steps to improve reliability. In this work, we use the sky-background-subtracted postage images with an area of 800 × 800 pixels.

Figure 1 shows the color image cutout from the SDSS skyserver (panel (a)) and the SDSS g-, r-, and i-band combined image of U141 in panel (b) with a size of $100^{\prime\prime} \times 100^{\prime\prime}$. Following the scheme and weight factors (Equation (1) in Kniazev et al. 2004) for each band, we combined the SDSS g-, r-, and i-band images to increase the signal-to-noise ratio by a factor of ∼$\sqrt{3}$ compared with that in the r-band image alone (Lisker et al. 2006b). Double cores and a boxy shape are clearly seen in Figures 1(b) and (c), respectively. In Figure 1(b), we show a contrast-stretched version to make the double cores more visible at the center of the galaxy. In Figure 1(c), where we overlaid the contours, three-pixel smoothing has been applied to enhance the signal at the outer part of the galaxy.

Basic information on U141 is summarized in Table 1. The brightness and the coordinates are adopted from Table 2 in Pak et al. (2014) and from the NASA/IPAC Extragalactic Database. Structural parameters are from surface photometry (see below). We calculate the total star formation rate (SFR) from the FUV flux measured in Pak et al. (2014) and using Equation (3) from Lee et al. (2009).

Table 1.  Basic Information on U141

Parameter	Information	Reference
Galaxy name	MCG+08-22-082, U141,	(1), (2)
	PGC 038825, SDSS J121122.56+501610.2	(1)
R.A.	1828448 (12h 11m 225, J2000)	(1)
decl.	+502696 (+50° 16' 10'', J2000)	(1)
l, b	138557908, 65625904	(1)
$u,g,r,i,z$	17.28 ± 0.04, 15.76 ± 0.02, 15.19 ± 0.02, 14.88 ± 0.02, 14.93 ± 0.03 mag	(2)
Mr	−16.01 mag	(2)
Re	1.7 kpc	(3)
${\mu }_{e}$	25.54 mag arcsec−2	(3)
Sérsic index, n	1.3	(3)
Redshift, z	0.003	(1)
Radial velocity, vr	899 km s−1	(1)
Total star formation rate, log SFRFUV (M☉ yr−1)	−2.85	(4)

References. (1) NASA/IPAC Extragalactic Database, (2) Pak et al. (2014), (3) this study (SDSS r-band image), (4) this study (using the GALEX FUV flux).

Download table as:  ASCIITypeset image

Table 2.  Results of Fits to the Surface Brightness Profiles

Band	${\mu }_{0}$	${\mu }_{e}$	Re	n
	(mag arcsec−2)	(mag arcsec−2)	(arcsec)
u	24.94 ± 0.12	27.17 ± 0.12	16.59 ± 1.64	1.18 ± 0.12
g	23.54 ± 0.02	26.23 ± 0.02	22.34 ± 0.25	1.40 ± 0.02
r	23.17 ± 0.02	25.54 ± 0.02	21.31 ± 0.22	1.25 ± 0.02
i	22.96 ± 0.02	24.87 ± 0.02	22.02 ± 0.27	1.23 ± 0.02
z	22.65 ± 0.06	25.25 ± 0.06	17.90 ± 0.92	1.20 ± 0.07

Download table as:  ASCIITypeset image

We derive a major axis light profile from fitting the elliptical isophotes in the images of all bands. The IRAF/ELLIPSE (Jedrzejewski 1987) task is used for this purpose, in which the sky-background-subtracted images were used and unrelated foreground and background objects were masked manually. During the ellipse fitting, we fix the center, while the position angle (PA) and the ellipticity of isophotes are allowed to vary. We provide the initial values of these isophotal parameters as returned from Source Extractor (SExtractor; Bertin & Arnouts 1996) photometry. We separately measure the surface brightness profiles in the images of all bands using the same initial parameters.

Figure 2 shows the results. We present the surface brightness profiles of U141 in all five filters in panel (a), in which we overplot u-, g-, r-, i-, and z-band surface brightness profiles from bottom to top. As expected, the r-band data points have the smallest error bars, and therefore we primarily use the r-band surface photometry to derive the structural parameters of U141.

Zoom In Zoom Out Reset image size

The observed surface brightness profile $\mu (R)$ is fitted with a Sérsic function of the form given in

$$\begin{eqnarray}&&\mu (R)={\mu }_{0}+\displaystyle \frac{2.5}{\mathrm{ln}(10)}{\left(\displaystyle \frac{R}{h}\right)}^{1/n},\end{eqnarray} \tag{ 1 }$$

where ${\mu }_{0}$ is central surface brightness, h is scale length, and n is the Sérsic index (Graham & Driver 2005). This process gives the values of ${\mu }_{0}$, h, and n as output. An efficient curve fitting code, MPFIT, which is implemented in the IDL library, has been used to obtain best-fit parameters in each band (Markwardt 2009). The inner $5^{\prime\prime}$ is excluded during the fit to avoid the central double cores.

The effective radius R_e is obtained using the equation

$$\begin{eqnarray}&&{R}_{e}={b}^{n}h,\end{eqnarray} \tag{ 2 }$$

where $b=1.9992n-0.3271$, for $0.5\lt n\lt 10$, and the effective surface brightness, ${\mu }_{e}$, is calculated using

$$\begin{eqnarray}&&{\mu }_{0}={\mu }_{e}-2.5b/\mathrm{ln}(10)\end{eqnarray} \tag{ 3 }$$

(Graham & Driver 2005).

All the derived structural parameters in each band are summarized in Table 2. Each column presents (1) band, (2) central surface brightness, (3) effective surface brightness, (4) effective radius, and (5) Sérsic index. The errors of surface brightness profiles are measured from the rms scatter of intensity along the ellipse fit. These errors are used as weight values to calculate the errors of the fitted Sérsic parameters using the MPFIT package, which measures errors from the covariance metrics. We find that U141 has a nearly exponential light profile, with the best-fit Sérsic index equal to 1.3 for the r-band. This is consistent with the results from other bands. The r-band effective radius is ${R}_{e}=21\buildrel{\prime\prime}\over{.}31$ (1.7 kpc),⁶ which is consistent with the results from the g- and i-bands, but is much larger compared to the values from the u- and z-bands. The effective surface brightness in the r-band is ${\mu }_{e,r}\;=25.54$ mag arcsec⁻².

For the purpose of obtaining the radial g − r color profile, we match the image quality in the g- and r-bands. FWHM estimates of the g- and r-bands are 13 and 12, respectively. We therefore degrade the r-band image by smoothing with a Gaussian kernel of 0.6 pixels. We consider the point between the cores as the photometric center. Finally, we derive the color distribution from the azimuthally averaged light profiles in the g- and r-bands. In Figure 2(b), we show the g − r color profile of U141; here we can clearly see that, as in the color image, a positive color gradient is eminent, where ${\rm{\Delta }}(g-r)$ is 0.18 mag within the R_e. The color profile becomes nearly flat beyond 10'', with an average g − r color index of 0.6 mag.

We show the variation of the isophotal parameters of PA, ellipticity (e), and boxiness parameter (${a}_{4}/a$) in the third, fourth, and fifth panels of Figure 2, respectively. All three parameters remain nearly unchanged beyond 10'' where the average values of PA and ellipticity are 79° and 0.5, respectively. The ELLIPSE task measures isophote deviations and gives amplitudes using a Fourier series shown in

$$\begin{eqnarray}&&I(\theta )={I}_{0}+{{\rm{\Sigma }}}_{n=1}^{\infty }({A}_{n}\mathrm{cos}n\theta +{B}_{n}\mathrm{sin}n\theta ),\end{eqnarray} \tag{ 4 }$$

where θ is the azimuthal angle and I₀ is the average intensity over the ellipse.

The A_n and B_n, given by the ELLIPSE task, are the higher-order Fourier coefficients divided by the semimajor axis length, a, and the local gradient along the major axis (Hao et al. 2006). As a boxiness parameter, ${a}_{4}/a$ is defined as $\sqrt{(1-e)}{A}_{4}$ (see Milvang-Jensen & Jorgensen 1999 and Hao et al. 2006 for details), where A₄ from the ELLIPSE task is a measure of the isophotal deviation from an ellipse.⁷ Negative and positive values of ${a}_{4}/a$ indicate boxy and disky shapes of a galaxy, respectively (Bender et al. 1988; Hao et al. 2006). In Figure 2 (e), ${a}_{4}/a$ of U141 shows a variation from −0.15 to 0.15. Interestingly, while the value of ${a}_{4}/a$ is mostly between 0 and $0.05$ within 10'', even reaching about −0.1 at the extreme, the average value of ${a}_{4}/a$ beyond 10'' radius is −0.06. This implies that the outer isophotes of U141 might show boxy shapes.

At first glance U141 is a typical blue-centered dE, as Lisker et al. (2006a) have identified in the Virgo cluster. But, interestingly, a careful inspection of the SDSS images shows that U141 possesses double cores and they are notably bluer than the main body of the galaxy (see Section 3.1 for a discussion on this in the context of the merger scenario).

To perform photometry and measure the stellar masses of these cores, we subtracted a smooth galaxy model that we had obtained from surface photometry in the previous section. In the SDSS r-band image, the centroids of the two cores are separated by 10 pixels, corresponding to a physical separation of 39 = 312 pc. While we consider the center of the fitted ellipses in the outer part (i.e., beyond 10'') to be the global photometric center of the galaxy, this does not coincide with the positions of either of the two cores.

We show the galaxy-subtracted residual image in Figure 3, in which the cores are named A (east) and B (west). We perform aperture photometry to measure the total flux. During the measurement of the flux of each core, we first mask the other core and use a circular aperture of 10 pixels. The sky background is selected from an annulus of 11 and 15 pixels as the inner and outer radii, respectively. We find that the two cores are similar in brightness, but core A is slightly bluer and more compact than B; we can also see this in the SDSS color image (Figure 1(a)). Using the formula provided by Bell et al. (2003), we derive rough stellar masses of these cores from the r-band luminosities and g − r colors. The results are listed in Table 3. Since the mass-to-light ratio obtained from Bell et al. (2003) heavily depends on color only, the redder core B seems more massive than core A. However, these mass estimates should be taken as rough estimates, not exact values.

Zoom In Zoom Out Reset image size

Table 3.  Properties of the Two Cores

Core	Mr	g − r	M*	SFR	12 + log(O/H)	vr
	(mag)	(mag)	(106 M☉)	(10−3 M☉)	(dex)	(km s−1)
A	−11.08	−0.22	0.5	0.7	8.4	$781\pm 2.6$
B	−10.97	0.10	1	0.1	8.6	$778\pm 8.8$

Download table as:  ASCIITypeset image

We find that the SDSS has targeted U141 twice, placing the fiber of each core in different epochs. The spectrum of B is already available in DR7, and later DR12 adds that of A. The SDSS spectra are obtained with fibers of 3'' diameter. We show the positions of fibers on U141 in Figure 3, where the crosses represent the positions and circles represent the fiber area. It seems that the fibers are fairly well-centered at the centroid of the cores. There is almost no overlap between them, but given the typical seeing of ∼1'' of SDSS observations, it is likely that a slight fraction of fluxes might have been exchanged.

The optical spectra of these cores exhibit the emission lines of typical H ii regions (Figure 4). The radial velocities of the A and B cores are $781\pm 2.6$ km s⁻¹ and $778\pm 8.8$ km s⁻¹, respectively. This difference is well within the uncertainty of the radial velocities as measured from the SDSS spectroscopy. The emission line fluxes are measured after subtracting the stellar absorption features using the publicly available IDL code GANDALF (Sarzi et al. 2006) and the stellar templates of Tremonti et al. (2004).

Zoom In Zoom Out Reset image size

Gas-phase oxygen abundances, 12 + log(O/H), for the two cores were estimated using the two methods described, among others, by Marino et al. (2013): the so-called N2 and O3N2 methods. The N2 method only considers the line ratio between Hα and [N ii], while the O3N2 method uses a combination of the line ratios Hα/[N ii] and [O iii]/Hβ. We obtained 12 + log(O/H) = 8.4(8.3) and 8.6(8.5) dex from the N2(O3N2) method for the A and B cores, respectively. The systematic error of these two methods is 0.2 dex. We measure the SFR from ${\rm{H}}\alpha$ emission flux and calibration provided by Kennicutt (1998). For the A and B cores, the SFRs are $0.7\times {10}^{-3}$ and $0.1\times {10}^{-3}$ M_☉ yr⁻¹, respectively. The sum of the SFRs in the two cores derived from Hα emissions is nearly half of the global value derived from total FUV emission. The Hα equivalent widths are 83 and 6 Å for the A and B cores, respectively.

Figure 5 presents the spatial distribution of galaxies in the Ursa Major cluster field and neighbors of U141 in order to show their environments. U141 is located in the middle of the cluster (Figure 5(a)) but not in the dense region. Within the sky projected radius of 100 kpc, there are only two close neighbors, UGC 7176 (U138) and NGC 4157 (U140). U141 has relative radial velocities of 131 and 207 km s⁻¹ to UGC 7176 and NGC 4157, respectively. NGC 4157 is a bright S0 galaxy and UGC 7176 is a dwarf irregular galaxy with magnitudes of ${M}_{r}=-20.30$ and −15.47 mag, respectively.

Zoom In Zoom Out Reset image size

We carefully inspect the SDSS r-band image of this region to search for any tidal features (i.e., stellar stream or filament) that may have originated from the past interaction among galaxies. Within the detection limit of the SDSS, no such feature is observed around either U141 or its neighboring galaxies (Kim et al. 2012; Kreckel et al. 2012; Hirschmann et al. 2013).

With the discovery of morphological diversity in dEs by Lisker et al. (2006a, 2006b), our understanding of the formation and evolution of these low-mass objects has become even more complicated. A considerable fraction of dEs possess substructural features (i.e., a blue center, a disk spiral arm, a bar, and/or a central nucleus), and such substructural varieties are inexplicable within a single formation scenario of dEs. The role of the environment, which can act in several different ways (see the reviews Boselli & Gavazzi 2006 and Lisker 2009), has been frequently discussed in the literature. Although the strong environmental effects could be thought of as the role for the various morphologies of dwarf galaxies, other channels for the morphological diversity in dEs can be suggested.

The star clusters or super star clusters in a galaxy can be observed as double cores, just like those embedded in U141. It is possible that the star clusters go into the centers of galaxies due to the dynamical friction and the sizes of the cores and star clusters might be similar. Mean effective radii of star clusters are 3−5 pc (0038–0063) at the distance of the Ursa Major cluster (Seth et al. 2006). This size cannot be resolved in our SDSS r-band image with a seeing of 12. In the r-band, the FWHMs of cores A and B are 81 and 43, respectively. The FWHMs of double cores are large enough to be resolved at the distance of the Ursa Major cluster, so these cores might not be star clusters or super star clusters.

Another possibility for the origin of the double cores is that they could be star-forming H ii regions. Gu et al. (2006) presented two distinct nuclei of IC 225 in the off-center region, which is somewhat similar to what we observed in U141. They suggested that the off-center nucleus originates from star-forming activity with high metallicity (12 + log(O/H) = 8.98). The metallicity measured in this region is higher than that (12 + log(O/H) = 8.2) calculated by using the luminosity-metallicity relation of isolated dwarf galaxies (Duc et al. 2004) at ${M}_{B}=-17.14$. In the case of U141 (${M}_{B}\sim -15.03$),⁸ the cores of U141 have higher metallicities (12 + log(O/H) = 8.4 and 8.6) than those (12 + log(O/H) ∼ 8.0) of isolated galaxies in the luminosity-metallicity relation (Duc et al. 2004). This implies that the cores might be star-forming H ii regions.

Debattista et al. (2006) found double nuclei with ages of ∼8 Gyr at the center of dwarf elliptical galaxy VCC 128, which has a absolute magnitude of ${M}_{B}=-15.5$. To maintain the observed configuration of double nuclei with a small separation of ∼32 pc, they introduced a supermassive black hole without any direct or indirect evidence. If the double nuclei are nuclear disks surrounding a supermassive black hole with $\sim {10}^{6}$ ${M}_{\odot }$, they should be in a stable, long-lived configuration and could account for the old populations. In the case of U141, the separation of the double cores is 312 pc. This is much larger than the Bondi radius (14 pc) of a black hole (Bondi 1952), where the matter is gravitationally bound to and is being accreted into the black hole, assuming that the mass of the black hole is $\sim {10}^{6}\;{M}_{\odot }$ in the middle of the double cores and assuming that there is a relative velocity of ∼25 km s⁻¹ between the black hole and the cores, which is the velocity dispersion of a galaxy with ${M}_{B}\sim -15.0$ (De Rijcke et al. 2005). The separation of the double cores in the central massive black hole hypothesis is too large to accrete matter into the central black hole. Therefore, U141 might not have a massive black hole.

A galaxy merger could be a possible mechanism for forming the double cores. Double cores are frequently observed in giant early-type galaxies with disturbed features (Sheen et al. 2012; Nesci et al. 2015). Dwarf galaxies with double cores are rarely reported. U141's circumstances are well-suited for a possible merger, i.e., it is in a low-density cluster with a low velocity dispersion.

Boxy isophote dEs are rarely observed. LEDA 074886 is an almost rectangular-shaped dE, while its surface brightness profile is exponential. Graham et al. (2012) suggested a hybrid formation scenario in which LEDA 074886 might have emerged through both a dissipational merger in the center and a dry merger in the outer part of the galaxy. Kazantzidis et al. (2011) performed simulations of the formation of dEs by merging two dwarf disk galaxies. According to their results, a boxy morphology of U141 could be formed by both major and minor mergers.

The fact that U141 shows a mixture of features with double cores and a boxy-shaped morphology in the Ursa Major cluster (low cluster velocity dispersion and no intracluster medium) suggests a dwarf-dwarf merger for the formation of U141. Although the double cores in U141 might be explained with star-forming H ii regions, the existence of the H ii regions does not simultaneously explain the boxy shapes. In the case of a merger of dwarf galaxies, double cores and the boxy shape of U141 can be simultaneously explained. Although it is expected that low-mass galaxies will experience less merging events (De Lucia et al. 2006), examples of dwarf-dwarf mergers have been multiplying significantly in recent studies (Martínez-Delgado et al. 2012; Amorisco et al. 2014; Paudel et al. 2015).

During the merger of galaxies, the gas supply toward the center becomes extremely efficient, producing a central starburst (Bekki 2008; Peeples et al. 2009). When the merger-induced central starburst is triggered, the galaxy will look like a blue-centered galaxy. In addition, it is expected that the central gas metallicity will be lower than that expected from the luminosity-metallicity relation, due to the pure gas infall (Peeples et al. 2009). SFR also will be larger than that expected from the SFR–mass relation (Peeples et al. 2009). However, U141 has higher gas metallicities (12 + log(O/H) = 8.4 and 8.6) for the two cores and a lower SFR (log SFR [${M}_{\odot }$ yr⁻¹] ∼ −2.85) than the expected metallicity (12 + log(O/H) ∼ 8.0) and the SFR estimated from the SFR–mass relation (log SFR ∼ −1.6), respectively. U141 is similar to the galaxies with high gas metallicity and the low SFR found by Peeples et al. (2008). They predicted that those galaxies are galaxies with depleted H i compared to the abundance of oxygen in the final stages of their star-forming activities. Within this context, we suspect that U141 might have been formed from a merger between H i-depleted galaxies with an overall low SFR in the end of their star formation activity. During the merger of these H i-depleted progenitors, the star formation efficiency of U141 might have been low and gas metallicity might have been high.

If U141 was formed by a merger, we can roughly estimate the timescale of a merger between two galaxies for which we assume a mass ratio of 2:1 between the primary and secondary based on the mass ratio of the two cores. We use Equation (4) from Jiang et al. (2008) for the dynamical friction timescale, where the secondary starts to merge from the virial radius of the primary with an elliptical orbit ( = 0.5) and a relative radial velocity of 158 km s⁻¹, which is the velocity dispersion of the Ursa Major cluster. The virial radius is converted from the effective radius using the relation between virial radius and effective radius (Kravtsov 2013). We obtain a merging timescale of ∼1 Gyr for the two progenitors to arrive at a separation of ∼312 pc and a relative radial velocity of ∼3 km s⁻¹. In the end, the cores will be merged within ∼250 Myr.

Nevertheless, the shallow SDSS images may prevent us from excluding other possible scenarios, particularly tidal interaction with other nearby galaxies. We cannot entirely rule out the possibility that U141 is at the stage of ongoing interaction with either UGC 7176 or NGC 4157. Given the very interesting nature of U141, we further suggest that a deep imaging and kinematic study would be extremely valuable.