Interacting galaxy NGC4656 and its unusual dwarf companion
Anatoly V. Zasov, Anna S. Saburova, Oleg V. Egorov, Roman I. Uklein
MMNRAS , 1–8 (2016) Preprint 21 September 2018 Compiled using MNRAS L A TEX style file v3.0
Interacting galaxy NGC4656 and its unusual dwarf companion (cid:63)
Anatoly V. Zasov , † , Anna S. Saburova , Oleg V. Egorov , Roman I. Uklein Sternberg Astronomical Institute, M.V. Lomonosov Moscow State University, Universitetskij pr., 13, Moscow, 119234, Russia Faculty of Physics, M.V. Lomonosov Moscow State University, Leninskie gory 1, Moscow, 119991, Russia Special Astrophysical Observatory, Russian Academy of Sciences, Nizhniy Arkhyz, Karachai-Cherkessian Republic 357147, Russia
21 September 2018
ABSTRACT
We studied the nearby edge-on galaxy NGC4656 and its dwarf low surface brightness com-panion with the enhanced UV brightness, NGC4656UV, belonging to the interacting systemNGC4631 /
56. Regular photometric structure and relatively big size of NGC4656UV allowsto consider this dwarf galaxy as a separate group member rather than a tidal dwarf. Spectrallong-slit observations were used to obtain the kinematical parameters and gas-phase metal-licity of NGC4656UV and NGC4656. Our rough estimate of the total dynamical mass ofNGC4656UV allowed us to conclude that this galaxy is the dark-matter dominated LSB dwarfor ultra di ff use galaxy. Young stellar population of NGC4656UV, as well as strong local non-circular gas motions in NGC4656 and the low oxygen gas abundance in the region of thisgalaxy adjacent to its dwarf companion, give evidence in favour of the accretion of metal-poor gas onto the discs of both galaxies. Key words: galaxies: individual: NGC4656, galaxies: kinematics and dynamics, galaxies:evolution, galaxies: abundances, galaxies: structure
Tidal interaction or merging of gas-rich galaxies often lead to thenascence of local sites of star formation in tidal structures in localregions of enhanced density of gas lost by galaxies. Under certainconditions gravitationally bound tidal dwarf galaxies (TDGs) maybe formed in the tidal debris (most often – in the tidal tails). Theirdistinguishing features are the young stellar population, a low, ifany, content of dark matter, and the moderate gas metallicity, be-cause they consist of gas expelled from the peripheral regions ofmore massive parent galaxies. Some non-bound and short livedsites of star formation may also appear beyond the main bodies ofgalaxies as a result of interaction. In addition, a close encounter orcollision of galaxies may inspire a gas outflow and / or gas exchangewhich in turn may influence the properties of interstellar mediumand gas abundance not only of the interacting galaxies, but also oftheir dwarf satellites.The current paper is devoted to the study of the late typegalaxy NGC4656, interacting wth NGC4631, and its unusual dwarfcompanion – NGC4656UV. A system NGC4631 /
56 consists of twomoderate-size interacting galaxies (Combes 1978) separated byabout a half degree (approximately 65 kpc), containing the regionsof intense star formation, evidently triggered by interaction. Bothgalaxies are late-type and gas-rich systems observed nearly edge- (cid:63)
Based on observations obtained with the 6-m telescope of the Special As-trophysical Observatory of the Russian Academy of Sciences (SAO RAS). † E-mail: [email protected] on. Observations of H i reveal a complex character of gas dynamicsand gas distribution in galaxies with the far outflowing extraplanarstreams of low density H i from NGC4631 (Rand 1994). The op-tical tidal stream and several low surface brightness dwarf galaxieswere also discovered in the vicinity of NGC4631 (Karachentsevet al. 2014; Martínez-Delgado et al. 2015).The second galaxy, NGC4656, is not so disturbed as its com-panion, although its surface brightness distribution is highly asym-metric, so that the NE-half contains the largest H ii regions beyondthe central area, and looks much brighter than the dim opposite half.UV images of this galaxy obtained by GALEX allowed to revealthe unusual satellite of about 10 kpc in diameter which is barelyseen in the SDSS maps (Schechtman-Rook & Hess 2012) (see Fig.1). This peculiar object – NGC4656UV — is a low surface bright-ness galaxy if to admit that it is a bound system. It is connectedwith NGC4656 by H i and probably stellar bridge. A systemic ve-locity of NGC4656UV is close to that for NGC4656, which leavesno doubt that this object is a part of the interacting system. Multi-wavelength archival data analysis carried out by Schechtman-Rook& Hess (2012) for NGC4656, led the authors to the conclusionsthat NGC4656UV is a low-metallicity tidal dwarf candidate, con-taining a relatively low amounts of dark matter, although the otherscenarios were also considered.The distance to NGC4656 and NGC4656UV is not wellknown. The brightest galaxies in the group – NGC4656 andNGC4631 – have systemic velocities with respect to Local Group c (cid:13) a r X i v : . [ a s t r o - ph . GA ] M a y A. Zasov et al. V LG =
636 and 665 km s − respectively (according to Hyperleda database, (Makarov et al. 2014)), corresponding to about 9 Mpcfrom the Hubble relation. The Updated Catalog of nearby galax-ies (Karachentsev et al. 2013) gives the distance 5.4 Mpc forNGC4656 (TF-method), although the direct distance estimate forNGC4631 (TRGB-method) gives 7.38 Mpc. Schechtman-Rook &Hess (2012) used the distance 7.2 Mpc, following Seth et al. (2005)(TRGB-method for NGC4656). We adopted this value for the sys-tem. An ultra-faint H α ¸Ssheet ˇT has been revealed between NGC4656 and NGC 4631 by Donahue et al. (1995). However, as it wasnoted in Schechtman-Rook & Hess (2012), NGC 4656UV is notdetected in these observations.The absence of spectral optical data makes di ffi cult to clar-ify the nature of NGC4656UV. In this paper we describe the at-tempt to measure the velocity and metallicity of emission gas ofNGC4656UV and NGC4656 using the long-slit mode of observa-tion at 6m telescope BTA of Special Astrophysical Observatory.The current paper is organized as follows: in Sect. 2 we de-scribe details and results of the data reduction, Sect. 3 is devoted tothe discussion and in Sect 4 we give the main conclusions. We observed NGC4656 and NGC4656UV in 2013-2016 withthe spectrographs SCORPIO-2 (Afanasiev & Moiseev 2011) andSCORPIO (Afanasiev & Moiseev 2005) at the prime focus of the 6-m Russian telescope BTA at Special Astrophysical Observatory ofthe Russian Academy of Sciences (SAO RAS). The dates of obser-vations, exposure times, seeing, observers and dispersers are givenin Table 1. In our observations we used two di ff erent dispersers:the grism VPHG1200@540 which covers the spectral range 3600-7070 Å and has a dispersion of 0.87 Å pixel − , spectral resolu-tion FWHM ≈ . / px and thespectral resolution of 2.2 Å. The scale along the slit is 0.36 (cid:48)(cid:48) / px, theslit width is 1 (cid:48)(cid:48) .The positions of the slit are shown in Fig. 1 (rightpanel).The data reduction was performed in a standard way using idl -based reduction pipeline. The full description of the reduction canbe found e.g. in Zasov et al. (2016) and Zasov et al. (2015). Shortly,the procedure included a bias subtraction and truncation of over-scan regions; a division by normalized flat field frames; the wave-length calibration using the spectrum of a He-Ne-Ar calibrationlamp; linearization and summation; the night sky subtraction; theflux calibration using the spectra of the standards HZ44 and Feige56 that were obtained during the same or adjacent nights. The vari-ation of instrumental profile of the spectrograph was measured andtaken into account. The sky-subtracted spectrum of NGC4656UVfor PA = ◦ , integrated along the slit is demonstrated in Fig. 2.We fitted the reduced and binned spectra with high-resolutionPEGASE.HR simple stellar population models (Le Borgne et al.2004) convolved with the variation of the instrumental profile usingNB ursts full spectral fitting technique (Chilingarian et al. 2007)(for more details see e.g. Zasov et al. 2016). From this fitting weget the parameters of stellar population: age and metallicity, line-of-sight velocity, velocity dispersion and Gauss-Hermite moments http: // leda.univ-lyon1.fr / h3 and h4 which characterize the deviation of LOSVD from theGaussian profile. In order to obtain the velocity and velocity dis-persion of the ionized gas as well as the fluxes of emission lines wesubtracted the stellar population model spectra from the observedones and fitted the emission lines by Gaussian distribution.The oxygen abundance, which is the indicator of gas metallic-ity, was estimated using the measured fluxes of emission lines. Wewere able to use ‘direct’ T e method only for several brightest re-gions of NGC4656 because of the faintness of sensitive to electrontemperature [O iii ] 4363 Å emission line. Therefore, we used sev-eral calibrations that are based on the strong emission lines. Thereis a large discrepancy (up to 0.5 dex) between the metallicity esti-mates obtained by di ff erent methods (see, e.g., Kewley & Ellison2008; López-Sánchez et al. 2012). We selected one empirical (cali-brated by the sample of H ii regions with well measured metallicityusing T e method) and one theoretical (calibrated using photoion-ization models with pre-defined metallicity) methods: S methodproposed by Pilyugin & Grebel (2016) and izi method from Blancet al. (2015), respectively. It is necessary to know the relative ra-tio of [O iii ] 5007 Å, [N ii ] 6584 Å and [S ii ] 6717,6731 Å toH β fluxes in order to apply S-method, while izi uses Bayesian in-ference to estimate oxygen abundance from all available emissionlines fluxes. Note that we do not utilize the R-method proposed byPilyugin & Grebel, because it uses [O ii ] 3727 Å line, which isvery noisy for significant part of our spectra (yet there is a goodagreement between the estimates obtained by S and R methods forthe area of bright [O ii ] line). We will refer further to the oxygenabundance estimated using S and izi methods as 12 + log (O / H) S and 12 + log (O / H) izi respectively. The enhanced UV brightness of NGC4656UV and numerouspatches observed mostly at its peripheric regions clearly demon-strate the presence of young stellar population. The structure ofthe dwarf is rather uneven. However, the r-band surface brightnessradial profile of NGC4656UV, that we obtained from SDSS data,follows the distribution law of exponential disc: µ r ( r ) = ( µ ) r + . r / h ) , (1)where ( µ ) r and h are the disc central surface brightness andthe exponential scalelength, correspondingly (see Fig. 3). Theline in the figure denotes the result of the fitting with ( µ ) r = .
12 mag arcsec − (non-corrected for disc inclination) and h = (cid:48)(cid:48) = . ∼ . · L (cid:12) ,which after correction for the axes ratio of the outer isophotes a / b ≈ L r , c ∼ . · L (cid:12) .A profile of rotational velocity of NGC4656UV is shown inFig.4. The points in the figure correspond to the weighted meanvelocities in the radial bins of ±
20 arcsec width. Zeroth radial co-ordinate was chosen as the centre of symmetry of r-image, whichis about 5 arcsec towards SE from the intersection point of the slits(see Fig.1). Emission lines with low S / N ratio were ignored. The ro-tation curve was obtained using the emission spectra for two slices:
MNRAS000
MNRAS000 , 1–8 (2016) nteracting galaxy NGC4656 and its dwarf companion +32 o m s m s m s h m s m s m s m s m s m s m s NGC4656UVNGC4656 +31 o m s h m s m s m s NGC4656UVNGC4656 +32 o m s m s m s h m s m s m s m s m s m s m s PA=98 o PA=35 o PA=28 o PA=57 o PA=65 o NGC4656UVNGC4656
Figure 1.
The images of NGC4656 and NGC4656UV from left to right: in u-band, FUV and NUV- bands and a gri- bands. The last one is shown withoverplotted positions of the slit. The major axis of the ellipse overlaid in the images is 7 kpc.
Table 1.
Log of observations Slit PA Date Exposure time Seeing Observers Disperser( ◦ ) (s) ( (cid:48)(cid:48) )NGC465635 08.02.2013 1800 1.5 Uklein, Katkov VPHG1200@54098 08.02.2013 2700 1.5 Uklein, Katkov VPHG1200@540NGC4656UV28 31.03.2016 5400 1.7 Uklein VPHG1200@54057 06.04.2016 2700 1.2 Uklein VPHG1200@54065 09.05.2016 8400 1.9 Uklein VPHG2300G R e l . F l u x PA=28 o [OIII] 5007[OIII] 4959H β α Figure 2.
The sky-subtracted spectrum of NGC4656UV for PA = ◦ , inte-grated along the slit. The flux is in units of 10 − erg / cm / sec / Å, the wave-length is in Å, (non-corrected for the redshift). PA = ◦ and PA = ◦ and the following equations: V ( r ) = V r ( r ) (cid:112) ( sec ( i ) − tan ( i ) cos ( φ ))sin( i ) cos( φ ) (2) r = r φ (cid:112) (sec ( i ) − tan ( i ) cos ( α )) (3) The positions of the slit are shown in Fig. 1, right panel. r , m ag ./ a r cs e c r, arcsec Figure 3.
Radial profile of surface brightness of NGC4656UV obtainedfrom SDSS-r data. The line corresponds to the exponential disc fit. No cor-rection for disc inclination is applied.MNRAS , 1–8 (2016)
A. Zasov et al.
Here r is the radial coordinate, V ( r ) is circular velocity and φ is theangle between radius-vector of a given point and the major axis ofa galaxy, r φ is the radial distance in the sky plane, V r is the line-of-sight velocity corrected for the systemic velocity (the approachingside is mirrored and averaged with the receding side) and i is incli-nation angle, which corresponds to the observed isophotes ratio a / b ∼
2. The systemic velocity of the galaxy is taken to be 590 km s − .The maximum velocity of rotation V = ±
10 km s − al-lows us to get the rough estimate of total dynamical mass of thegalaxy inside of radius r = = (cid:48)(cid:48) = M ≈ V · r / G ≈ . + . , − · M (cid:12) . This value is in good agreement with theresult of Schechtman-Rook & Hess (2012) obtained from the PVdiagram of H i . It corresponds to the dynamical mass-to-light ratio M dyn / L r ∼ − M (cid:12) / L (cid:12) . A blue colour of this galaxy (accord-ing to Schechtman-Rook & Hess 2012, ( g − r ) ≈
0) correspondsto M ∗ / L r ≈ M (cid:12) / L (cid:12) for di ff erent stellar population mod-els (see f.e. Into & Portinari 2013; Roediger & Courteau 2015).Hence the input of stellar mass into the total mass of the dwarf islow. The mass of gas prevails over the mass of stars: H i mass ofNGC 4656UV (estimated as 3 . · M (cid:12) by Schechtman-Rook &Hess (2012)), corresponds to the total mass of gas (including he-lium) ∼ · M (cid:12) . Even if we assume that most of the observedgas connected with this galaxy lays within the optical borders ofNGC 4656UV, although this is a clear exaggeration, and keep inmind that the gas mass is underestimated due to the unaccountedmass of the molecular gas (which is expected to be quite low sincethe metallicity of the gas is low), the conclusion is unavoidable thatthe observed baryonic mass is several times lower than the dynamicmass in this galaxy. Thus it is unlikely that NGC4656UV is a tidaldwarf, which should contain no dark matter (unless one supposesthat NGC 4656UV has large amount of dark gas (i.e. cold gas non-detected by its emission, see, e.g. Kasparova et al. 2014)). Anotherargument against the tidal formation of the galaxy is its large sizewhich is only 2-3 times lower than that of NGC 4656 parallel withthe exponential shape of its surface brightness radial profile whichevidences that NGC 4656UV is rather relaxed system. Most prob-ably we have a deal with the dark matter dominated LSB dwarfgalaxy. One should keep in mind, however, that the total dynamicalmass estimate of NGC4656UV remains uncertain and needs moreelaborate investigation.Because of the weakness of even strong emission lines in thespectra of NGC 4656UV, we were able to measure the oxygenabundance for only one H ii region, which reveals the brightestemission in [O iii ] 5007 Å, [N ii ] 6584 Å and [S ii ] 6717,6731Å among all studied areas in this galaxy. This region is locatedat the south-west edge of the galaxy and crossed by the slit withPA = ◦ (see Fig. 1). We got the following oxygen abundances esti-mates for it: 12 + log (O / H) S = . ± .
10 and 12 + log (O / H) izi = . ± .
12. As we already noted in Section 2, the discrepancy be-tween these estimates is a very common problem: theoretical meth-ods (like izi ) usually yield values higher (up to 0.5 dex) than em-pirical ones. Our estimates of oxygen abundance in NGC4656UVconfirm the findings of Schechtman-Rook & Hess (2012), who sup-posed that it should contain very low metallicity gas in order toexplain the lack of IR emission.
In Figs. 5, 6 we demonstrate the radial profiles of the velocity andvelocity dispersion for PA = ◦ and PA = ◦ respectively. The circlescorrespond to the ionized gas, asterisks denote the estimates for Figure 4.
The rotation velocities of ionized gas of NGC4656 obtained fromtwo spectral slices PA = ◦ and PA = ◦ . The weighted average value of ro-tation velocity from the approaching and receding sides is given for thelarge bins. Dotted line shows the rotation velocity amplitude of 40 km s − . ∆ X , PA = 35 ◦ v L O S , k m s − H α H β [OIII][SII]stars σ L O S , k m s − H α H β [OIII][SII]stars
120 90 60 30 0 30 60 90 120 150 R , arcsec7.27.57.88.18.48.7 + l og ( O / H ) , d e x SW NE iziST e Figure 5.
Radial distribution of line-of-sight velocity, velocity dispersionand oxygen abundance of NGC4656 along the slit PA = ◦ . Top panel cor-responds to the reference SDSS gri composite image. The zero point forthe velocity V sys =
646 km s − . Circles at the bottom panel correspondto the values obtained for each bin along the slit, while squares representthe values measured by integrated spectra of H ii regions listed in Table 2.Zero-point for radial distance is arbitrary chosen. stellar population. For PA = ◦ the parameters of stars appeared tobe very uncertain, thus we decided not to put them into the graphs. MNRAS000
646 km s − . Circles at the bottom panel correspondto the values obtained for each bin along the slit, while squares representthe values measured by integrated spectra of H ii regions listed in Table 2.Zero-point for radial distance is arbitrary chosen. stellar population. For PA = ◦ the parameters of stars appeared tobe very uncertain, thus we decided not to put them into the graphs. MNRAS000 , 1–8 (2016) n t e r a c ti ngga l a xy N G C it s d w a r f c o m pan i on Table 2.
Results of spectroscopy of H ii regions in NGC4656UV and NGC4656 galaxies Reg. Pos., F(H β ), [O ii ]3727 H γ [O iii ]4363 [O iii ]5007 H α [N ii ]6584 [S ii ]6717 [S ii ]6731 c(H β ) 12 + log(O / H)S 12 + log(O / H)izi 12 + log(O / H)Tearcsec 10 −
16 ergs − − NGC4656UV PA28 ÷
62 2 . ± . − − − . ± .
011 2 . ± .
430 0 . ± .
006 0 . ± .
005 0 . ± .
009 0 .
47 7 . ± .
10 7 . ± . − NGC4656 PA35 − ÷ −
101 7 . ± . − − − . ± .
088 2 . ± .
117 0 . ± .
024 0 . ± .
023 0 . ± .
046 0 .
00 7 . ± .
11 8 . ± . − − ÷ −
96 13 . ± . − − − . ± .
098 2 . ± .
945 0 . ± .
017 0 . ± .
021 0 . ± .
028 0 .
13 8 . ± .
10 8 . ± . − − ÷ −
92 17 . ± . − . ± . − . ± .
071 2 . ± .
500 0 . ± .
013 0 . ± .
015 0 . ± .
024 0 .
14 8 . ± .
10 8 . ± . − − ÷ −
82 200 . ± .
77 3 . ± .
072 0 . ± . − . ± .
015 2 . ± .
102 0 . ± .
002 0 . ± .
002 0 . ± .
004 0 .
30 8 . ± .
13 8 . ± . − − ÷ −
77 728 . ± .
16 1 . ± .
033 0 . ± .
004 0 . ± .
010 6 . ± .
010 2 . ± .
080 0 . ± .
001 0 . ± .
001 0 . ± .
002 0 .
55 8 . ± .
15 8 . ± .
05 8 . ± . − ÷ −
70 210 . ± .
72 3 . ± .
064 0 . ± . − . ± .
011 2 . ± .
103 0 . ± .
002 0 . ± .
002 0 . ± .
004 0 .
22 8 . ± .
14 8 . ± . − − ÷ −
66 83 . ± .
55 5 . ± .
127 0 . ± . − . ± .
017 2 . ± .
143 0 . ± .
004 0 . ± .
005 0 . ± .
008 0 .
23 8 . ± .
10 8 . ± . − − ÷ −
57 259 . ± .
82 3 . ± .
068 0 . ± . − . ± .
007 2 . ± .
067 0 . ± .
002 0 . ± .
002 0 . ± .
003 0 .
24 8 . ± .
10 8 . ± . − − ÷ −
48 35 . ± . − . ± . − . ± .
026 2 . ± .
213 0 . ± .
009 0 . ± .
009 0 . ± .
013 0 .
00 8 . ± .
10 8 . ± . − − ÷ −
42 26 . ± .
36 4 . ± .
233 0 . ± . − . ± .
029 2 . ± .
257 0 . ± .
008 0 . ± .
010 0 . ± .
017 0 .
00 8 . ± .
10 8 . ± . − − ÷ −
36 19 . ± . − . ± . − . ± .
037 2 . ± .
326 0 . ± .
012 0 . ± .
012 0 . ± .
017 0 .
00 8 . ± .
10 8 . ± . − − ÷ −
30 25 . ± . − . ± . − . ± .
038 2 . ± .
300 0 . ± .
011 0 . ± .
010 0 . ± .
018 0 .
18 8 . ± .
10 8 . ± . − − ÷ −
19 172 . ± .
49 3 . ± .
058 0 . ± . − . ± .
008 2 . ± .
083 0 . ± .
002 0 . ± .
002 0 . ± .
003 0 .
06 7 . ± .
16 8 . ± . − − ÷ − . ± . − − − . ± .
181 2 . ± .
383 0 . ± . − − . − . ± . −
15 75 ÷
84 343 . ± .
49 2 . ± .
036 0 . ± .
003 0 . ± .
010 3 . ± .
005 2 . ± .
083 0 . ± .
001 0 . ± .
001 0 . ± .
002 0 .
10 7 . ± .
13 8 . ± .
09 7 . ± . ÷
88 24 . ± . − . ± . − . ± .
019 2 . ± .
431 0 . ± .
006 0 . ± .
005 0 . ± .
009 0 .
00 7 . ± .
10 8 . ± . −
17 89 ÷
94 30 . ± . − . ± . − . ± .
019 2 . ± .
626 0 . ± .
006 0 . ± .
006 0 . ± .
009 0 .
01 7 . ± .
10 8 . ± . −
18 95 ÷
104 380 . ± .
58 1 . ± .
041 0 . ± .
004 0 . ± .
011 3 . ± .
006 2 . ± .
097 0 . ± .
001 0 . ± .
001 0 . ± .
002 0 .
17 7 . ± .
14 8 . ± .
12 7 . ± . ÷
112 65 . ± .
29 3 . ± .
125 0 . ± . − . ± .
013 2 . ± .
281 0 . ± .
003 0 . ± .
003 0 . ± .
005 0 .
00 7 . ± .
25 8 . ± . −
20 114 ÷
118 8 . ± . − − − . ± .
036 2 . ± .
132 0 . ± .
019 0 . ± .
020 0 . ± .
027 0 .
00 7 . ± .
22 8 . ± . − NGC4656 PA98 − ÷ −
63 1 . ± . − − − . ± .
029 2 . ± .
154 0 . ± .
023 0 . ± .
024 0 . ± .
032 0 .
00 7 . ± .
11 8 . ± . − − ÷ −
51 2 . ± . − − − . ± .
026 2 . ± .
176 0 . ± .
021 0 . ± .
021 0 . ± .
034 0 .
07 7 . ± .
11 8 . ± . − − ÷ −
44 4 . ± . − − − . ± .
031 2 . ± .
982 0 . ± .
011 0 . ± .
011 0 . ± .
020 0 .
09 7 . ± .
11 8 . ± . − − ÷ −
37 7 . ± . − − − . ± .
017 2 . ± .
533 0 . ± .
008 0 . ± .
008 0 . ± .
015 0 .
21 7 . ± .
10 8 . ± . − − ÷ −
29 45 . ± .
05 2 . ± .
039 0 . ± .
003 0 . ± .
007 3 . ± .
004 2 . ± .
084 0 . ± .
001 0 . ± .
001 0 . ± .
002 0 .
00 7 . ± .
13 8 . ± .
09 7 . ± . − ÷ −
22 120 . ± .
11 2 . ± .
035 0 . ± .
002 0 . ± .
006 4 . ± .
004 2 . ± .
067 0 . ± .
001 0 . ± .
001 0 . ± .
001 0 .
27 7 . ± .
14 8 . ± .
07 7 . ± . − ÷ −
16 283 . ± .
11 1 . ± .
012 0 . ± .
001 0 . ± .
002 3 . ± .
001 2 . ± .
031 0 . ± .
001 0 . ± .
000 0 . ± .
001 0 .
13 7 . ± .
10 8 . ± .
09 7 . ± . − ÷ − . ± .
07 2 . ± .
050 0 . ± . − . ± .
003 2 . ± .
083 0 . ± .
001 0 . ± .
001 0 . ± .
002 0 .
05 7 . ± .
12 8 . ± . − − ÷ − . ± .
07 2 . ± . − − . ± .
005 2 . ± .
191 0 . ± .
003 0 . ± .
003 0 . ± .
005 0 .
17 7 . ± .
21 8 . ± . −
30 0 ÷ . ± .
07 3 . ± .
100 0 . ± . − . ± .
003 2 . ± .
092 0 . ± .
001 0 . ± .
001 0 . ± .
002 0 .
18 7 . ± .
16 8 . ± . −
31 5 ÷
10 137 . ± .
11 2 . ± .
031 0 . ± .
002 0 . ± .
004 2 . ± .
002 2 . ± .
048 0 . ± .
001 0 . ± .
001 0 . ± .
001 0 .
28 7 . ± .
12 8 . ± .
12 7 . ± . ÷
15 41 . ± .
06 3 . ± .
065 0 . ± . − . ± .
003 2 . ± .
082 0 . ± .
001 0 . ± .
001 0 . ± .
002 0 .
10 7 . ± .
16 8 . ± . −
33 17 ÷
23 17 . ± . − . ± . − . ± .
006 2 . ± .
201 0 . ± .
003 0 . ± .
003 0 . ± .
005 0 .
13 7 . ± .
10 8 . ± . −
34 24 ÷
32 9 . ± . − . ± . − . ± .
009 2 . ± .
378 0 . ± .
005 0 . ± .
005 0 . ± .
009 0 .
00 7 . ± .
10 8 . ± . −
35 38 ÷
49 39 . ± . − . ± . − . ± .
004 2 . ± .
149 0 . ± .
002 0 . ± .
002 0 . ± .
003 0 .
17 7 . ± .
10 8 . ± . −
36 51 ÷
56 58 . ± .
07 2 . ± .
052 0 . ± .
003 0 . ± .
005 1 . ± .
003 2 . ± .
064 0 . ± .
001 0 . ± .
001 0 . ± .
001 0 .
11 7 . ± .
15 8 . ± .
12 7 . ± . ÷
63 14 . ± . − . ± . − . ± .
004 2 . ± .
250 0 . ± .
003 0 . ± .
003 0 . ± .
005 0 .
00 7 . ± .
10 7 . ± . − M N R A S , ( ) A. Zasov et al. ∆ X , PA = 98 ◦ v L O S , k m s − H α H β [OIII][SII] σ L O S , k m s − H α H β [OIII][SII]
80 60 40 20 0 20 40 60 80 100 R , arcsec7.27.57.88.18.48.7 + l og ( O / H ) , d e x NW NE iziST e Figure 6.
The same as in Fig. 5, but for PA = ◦ . Velocity profile along PA = ◦ reveals a velocity gradient alongthe galaxy, evidently reflecting the disc rotation and large non-circular motions and unusually high velocity dispersion locally ex-ceeding 100 km s − . Non-circular gas motions are also clearly seenalong PA = ◦ , passing through the NE part of galaxy, althoughthe velocity wiggles are not so large as in the central H ii regionscrossed by PA = ◦ . The restricted radial extension of our measure-ments for NGC4656 and non-circular velocities of emission gas weobserve do not allow us to obtain the reliable value of rotational ve-locity amplitude for the galaxy. However, H i data which are muchmore extended enabled to determine the rotational velocity of ∼ − Schechtman-Rook & Hess (2012). It corresponds to the to-tal dynamical mass of NGC4656 inside of 6 − M tot ≈ · M (cid:12) .In Figs. 5 and 6 we show also the distribution of oxygen abun-dance along the slits. Parallel with the oxygen abundance distribu-tion for each pixel along the slit (shown by circles), we give themean values obtained from the integrated spectra of individual H ii regions or areas of di ff use emission (shown by squares). The re-sults of the emission lines measuring and the metallicity estimatesobtained for these selected regions are listed in Table 2.In Fig. 7 we show the so-called BPT diagnostic diagrams(Baldwin et al. 1981) of emission lines flux ratios constructed forboth slits. All studied regions fall in the area corresponding to thephotoionization mechanism of excitation (under the black separa-tion line from Kewley et al. (2001)). It justifies the use of the em-pirical methods to estimate oxygen abundance. −2.0 −1.5 −1.0 −0.5 0.0−0.4−0.20.00.20.40.60.81.0−2.0 −1.5 −1.0 −0.5 0.0log([NII]6548,6584/H α )−0.4−0.20.00.20.40.60.81.0 l og ( [ O III] / H β ) −2.0 −1.5 −1.0 −0.5 0.0log([SII]6717,6731/H α )−0.4−0.20.00.20.40.60.81.0 l og ( [ O III] / H β ) Figure 7.
Diagnostic BPT diagrams [O iii ] / H β vs [S ii ] / H α and [N ii ] / H α constructed for each pixel along the slits (shown by circles) and for individ-ual H ii regions from Table 2 (shown by crosses). Solid lines, according tothe models of Kewley et al. (2001) and Kau ff mann et al. (2003), separateregions of pure photoionization excitation caused by young massive stars(below the black line), shock excitation (or another mechanism not causedby star formation; above both black and grey lines) and combined contri-bution of both mechanisms (between grey and black lines in the lefthandpanel). As it is seen from bottom panels of Figs. 5, 6, gas metallicityis higher at the southern (dimmest) part of NGC4656 galaxy andlower at the NE side of NGC4656, facing the UV satellite. For sev-eral brightest regions we were able to estimate oxygen abundanceusing T e method. The inferred values are in good agreement withestimates made with S method.As it follows from Figs. 5 and 6, a gas is not chemically ho-mogeneous in the NGC4656 galaxy. In Fig. 8 we show the oxy-gen abundance distribution along the deprojected galactocentricdistance normalized to R = . = ◦ (HYPERLEDA database), i = ◦ Schechtman-Rook & Hess 2012 respectively. In Fig. 8 onemay clearly see oxygen abundance gradient in NGC4656. This gra-dient found by izi-method (see Fig. 8, below) is in a good agree-ment with the estimate of the O / H gradient from Pilyugin et al.(2014) (grad O / H = / kpc), who used the bright emissionlines method.An intriguing feature is seen at the northern part of the galaxy,facing the UV dwarf ( ∼
90 arcsec along the slit PA =
35, seeFig.5). Both methods used for metallicity estimation show a dropof metallicity there: from 12 + log (O / H) S = . ± .
13 and12 + log (O / H) izi = . ± .
14 to 12 + log (O / H) S = . ± .
09 and12 + log (O / H) izi = . ± .
08, which approaches to the metallicityof UV dwarf (12 + log (O / H) izi = . − − , which also could be caused by the fallingexternal gas. Note, however, that this kinematic feature is slightlyshifted from the region of metallicity’ minimum. Noteworthy thatthis site of possible gas accretion coincides with the region of thevisual ‘bend’ of galactic body. Two galaxies - NGC4656 and its UV satellite are in a processof tidal interaction. A mean density, which is proportional to
MNRAS000
MNRAS000 , 1–8 (2016) nteracting galaxy NGC4656 and its dwarf companion + l og ( O / H ) [ S ] PA=35PA=9812+log(O/H)=8.15−0.06/kpc0.0 0.5 1.0 1.5 2.0 2.5 3.07.67.88.08.28.48.60.0 0.5 1.0 1.5 2.0 2.5 3.0R/R + l og ( O / H ) [ i z i ] PA=35PA=9812+log(O/H)=8.31−0.03/kpc
Figure 8.
The variation of oxygen abundance with the galactocentric dis-tance. The deprojected distance estimates are made using the morpho-logical parameters of NGC 4656 (PA = ◦ and R = .
93 kpc) takenfrom HYPERLEDA database. The inclination of the galaxy was adoptedas i = ◦ according to Schechtman-Rook & Hess (2012). M dyn / R opt , for dwarf satellite is comparable with, if not higher thanthe density of the main galaxy, hence the satellite cannot be de-stroyed by tidal forces during the close encounter of galaxies. Nev-ertheless the tidal forces between these galaxies are responsible forformation of a short bridge connecting them. From Schechtman-Rook & Hess (2012) (Fig 8) it follows that the line-of-sight velocityalong the H i - bridge changes at about 30 km s − within the dis-tance range of about 4 (cid:48) . It corresponds to about 3.6 km s − / kpc forthe assumed distance 7.2 Mpc. At the same time, our data give forthe dwarf galaxy, which lies at the extension of the bridge, a veloc-ity gradient (30 ± − / (cid:48)(cid:48) , or about (8 . ± .
5) km s − / kpc.It definitely exceeds the velocity gradient along the bridge, whichevidences that the UV dwarf is not a part of the tidal feature comingout of the galaxy. We conclude that NGC4656UV is rather LSB-dwarf galaxy rich of gas and of dark matter, with the recently en-hanced star formation responsible for its blue colour. As it was ar-gued above, star formation is most probably the result of accretionof low metal abundant gas onto dwarf satellite and the NE-part ofNGC4656.A possible source of accreting gas is the neighbor edge-ongalaxy NGC4631, situated at a distance of about 60 kpc in thesky plane. Fingers of H i emerging from NGC4631 including inthe direction of NGC4656 are very prominent in the H i distri-bution map around this galaxy (Rand 1994; Schechtman-Rook &Hess 2012). A total mass of H i in the most prominent tidal spursreaches 3 · M (cid:12) , which is almost half of the total mass of H i inside of NGC4631. Note that NGC4631 is the moderately under- abundant galaxy: according to Pilyugin et al. (2014) , the centralvalue of oxygen abundance 12 + log (O / H) S is 8.39 ± R . Hence, to explain the O / Habundance of accreted gas one can conclude that this gas was ini-tially situated beyond the optical radius of the galaxy. Taking intoaccount the gradient of O / H, such metal poor gas should come outof the outermost parts of the galaxy ∼ . R , which makes this ori-gin of the accretion less realistic. Rather, if the gas had previouslybelonged to NGC4631, it was diluted by an intergalactic medium.However, similar role may play the accretion of low-enriched gasfrom the extended gaseous discs of both NGC4656 and its satellite.Gas, which envelops the stellar bodies of these galaxies, is clearlyseen in the H i maps.Note that NGC4656UV strongly resembles H i -rich ultra-di ff use galaxies (UDGs) where star formation of low e ffi ciencytakes place, see Leisman et al. (2017). A very similar object of thiskind is the nearest galaxy of this type UGC2172, which is twiceas distant as NGC4656UV. This starforming UDG was recentlystudied in details by Trujillo et al. (2017). Both UGC2172 andNGC4656UV have similar central brightness ∼ mag / arcsec ,close stellar masses ∼ (1-4) · M (cid:12) and gas content M HI ∼ (2 − · M (cid:12) .In Fig. 9 we compare NGC4656UV by mass of H i and O / Hratio with gas-rich dwarf LSB galaxies of similar luminosities andcentral surface brightnesses taken from the sample of van Zee et al.(1997). The position of NGC4656UV is shown by star, square sym-bols correspond to the data from van Zee et al. (1997). We passedto B-band luminosity from L r luminosity and colour indices usingthe transformation equations. We took the estimate of the H i massfrom Schechtman-Rook & Hess (2012). The data of van Zee et al.(1997) were adjusted to the of Hubble constant h =
75 km s − / M pc .From the diagrams one can see that NGC4656UV does not outstandfrom the other LSB dwarfs.However, there are at least three peculiarities of this galaxywhich make it di ff erent from the majority of the other LSB dwarfs,possessing similar luminosity and gas content. First, this galaxyis currently interacting with its more massive neighbour, which ex-plains the non-circular gas motions, as well as the recent star forma-tion triggered by interaction and gas accretion. Second, star forma-tion in this galaxy reveals itself by non-resolved emission patchesscattered allover the disc in the absence of the extended H ii com-plexes, usually observed in Irr-galaxies. It evidences a low densityof gas with no large-scale instabilities, when most of UV radiationleaks out from the gas layer. Similar character of star formation isalso often observed in the far outskirts of discs of gas-rich galaxies(see f.e. Werk et al. 2010). Third, NGC4656UV is unusually bluewith respect to other dwarf galaxies, including UDGs: its color andspectral energy distribution corresponds to the luminosity-weightedage of low-abundant stellar population which does not exceed sev-eral hundred million years, although the presence of the old pop-ulation cannot be excluded (Schechtman-Rook & Hess 2012). Ev-idently, in the absence of recent star formation this small galaxywould hardly be noticeable at all in the optical bands. We studied the kinematics and oxygen abundance of galaxyNGC4656 and its neighbour UV-bright LSB dwarf galaxy Pilyugin et al. (2014) estimated the oxygen abundance in NGC 4631 us-ing C-calibration, which is consistent with the S-method used in our paper.MNRAS , 1–8 (2016)
A. Zasov et al. van Zee et al. (1997) L B , L s un M HI /L B , M sun /L sun NGC4656UV 7,4 7,6 7,8 8,0 8,2 8,41E71E81E9 NGC4656UV van Zee et al. (1997) L B , L s un Figure 9.
A comparison of NGC4656UV with the other LSB dwarf galaxies from van Zee et al (1997).
NGC4656UV, based on the long-slit spectral observations. Our es-timates of gas velocities of NGC4656UV parallel with photomet-rical and H i data speak in favour of this system to be the darkmatter-dominated gravitationally bound galaxy with the low sur-face brightness (( µ ) r = . mag / arcsec , non-corrected for discinclination), rather than a tidal dwarf candidate. The parameters ofNGC4656UV are close to that observed for gas-rich dwarf LSB-galaxies and ultra-di ff use galaxies. Note that in the absence of re-cent (or current) star formation it would be extremely hard to dis-cover such a low brightness object.Oxygen gas-phase abundances found for the brightest H ii -region of NGC4656UV as well as for the emission gas of the maingalaxy NGC4656 at the side facing UV dwarf, are low: 12 + log O / H = izi -method). Both the low abundance and non-circulargas motions in NGC4656 parallel with the observed young stellarpopulation of its dwarf satellite NGC4656UV are considered as theevidences of the current accretion of metal-poor gas on the discs ofboth galaxies as the result of tidal interaction. ACKNOWLEDGEMENTS
The authors thank the anonymous referee for valuable commentsthat helped to improve the paper. The authors acknowledge thesupport of RFBR grants 14-22-03006, 15-52-15050 (observations)and the Russian Science Foundation (RSCF) grant 14-22-00041(data processing and analysis). This research has made use of theLyon Extragalactic Database (LEDA, http: // leda.univ-lyon1.fr). Inthis study, we used the SDSS DR13 data. Funding for the SDSSand SDSS-II has been pro- vided by the Alfred P. Sloan Founda-tion, the Participating Insti- tutions, the National Science Founda-tion, the U.S. Department of Energy, the National Aeronautics andSpace Administration, the Japanese Monbukagakusho, the MaxPlanck Society, and the Higher Education Funding Council forEngland. The SDSS Web site is http: // / . The Rus-sian 6-m telescope is exploited under the financial support by theRussian Federation Ministry of Education and Science (agreementNo14.619.21.0004, project ID RFMEFI61914X0004). REFERENCES
Afanasiev V. L., Moiseev A. V., 2005, Astronomy Letters, 31, 194 Afanasiev V. L., Moiseev A. V., 2011, Baltic Astronomy, 20, 363Baldwin J. A., Phillips M. M., Terlevich R., 1981, PASP, 93, 5Blanc G. A., Kewley L., Vogt F. P. A., Dopita M. A., 2015, ApJ, 798, 99Chilingarian I. V., Prugniel P., Sil’Chenko O. K., Afanasiev V. L., 2007,MNRAS, 376, 1033Combes F., 1978, A&A, 65, 47Donahue M., Aldering G., Stocke J. T., 1995, ApJ, 450, L45Into T., Portinari L., 2013, MNRAS, 430, 2715Karachentsev I. D., Makarov D. I., Kaisina E. I., 2013, AJ, 145, 101Karachentsev I. D., Bautzmann D., Neyer F., Polzl R., Riepe P., Zilch T.,Mattern B., 2014, preprint, ( arXiv:1401.2719 )Kasparova A. V., Saburova A. S., Katkov I. Y., Chilingarian I. V., BizyaevD. V., 2014, MNRAS, 437, 3072Kau ff mann G., et al., 2003, MNRAS, 346, 1055Kewley L. J., Ellison S. L., 2008, ApJ, 681, 1183Kewley L. J., Dopita M. A., Sutherland R. S., Heisler C. A., Trevena J.,2001, ApJ, 556, 121Le Borgne D., Rocca-Volmerange B., Prugniel P., Lançon A., Fioc M.,Soubiran C., 2004, A&A, 425, 881Leisman L., et al., 2017, preprint, ( arXiv:1703.05293 )López-Sánchez Á. R., Dopita M. A., Kewley L. J., Zahid H. J., NichollsD. C., Scharwächter J., 2012, MNRAS, 426, 2630Makarov D., Prugniel P., Terekhova N., Courtois H., Vauglin I., 2014, A&A,570, A13Martínez-Delgado D., D’Onghia E., Chonis T. S., Beaton R. L., Teuwen K.,GaBany R. J., Grebel E. K., Morales G., 2015, AJ, 150, 116Pilyugin L. S., Grebel E. K., 2016, MNRAS, 457, 3678Pilyugin L. S., Grebel E. K., Kniazev A. Y., 2014, AJ, 147, 131Rand R. J., 1994, A&A, 285, 833Roediger J. C., Courteau S., 2015, MNRAS, 452, 3209Schechtman-Rook A., Hess K. M., 2012, ApJ, 750, 171Seth A. C., Dalcanton J. J., de Jong R. S., 2005, AJ, 129, 1331Trujillo I., Roman J., Filho M., Sánchez Almeida J., 2017, ApJ, 836, 191Werk J. K., et al., 2010, AJ, 139, 279Zasov A., Saburova A., Katkov I., Egorov O., Afanasiev V., 2015, MNRAS,449, 1605Zasov A. V., Saburova A. S., Egorov O. V., Afanasiev V. L., 2016, MNRAS,462, 3419van Zee L., Haynes M. P., Salzer J. J., 1997, AJ, 114, 2497MNRAS000