Resolving the Discrepancy of Distance to M60, a Giant Elliptical Galaxy in Virgo
aa r X i v : . [ a s t r o - ph . GA ] M a y Draft version September 30, 2018
Preprint typeset using L A TEX style emulateapj v. 12/16/11
RESOLVING THE DISCREPANCY OF DISTANCE TO M60,A GIANT ELLIPTICAL GALAXY IN VIRGO
Myung Gyoon Lee and In Sung Jang , Astronomy Program, Department of Physics and Astronomy, Seoul National University, Gwanak-gu, Seoul 151-742, Korea and Leibniz-Institut f¨ur Astrophysik Potsdam (AIP) An der Sternwarte 16, 14482 Potsdam, Germany
Draft version September 30, 2018
ABSTRACTThere is a well-known discrepancy in the distance estimation of M60, a giant elliptical galaxy inVirgo: the planetary nebula luminosity function (PNLF) distance moduli for this galaxy are, onaverage, 0.4 mag smaller than the values based on the surface brightness fluctuation (SBF) in theliterature. We present photometry of the resolved stars in an outer field of M60 based on deep F775Wand F850LP images in the Hubble Space Telescope obtained as part of the Pure Parallel Program inthe archive. Detected stars are mostly old red giants in the halo of M60. With this photometry, wedetermine a distance to M60 using the tip of the red giant branch (TRGB). A TRGB is detected atF850LP
TRGB = 26 . ± .
06 mag, in the luminosity function of the red giants. This value correspondsto F814W , TRGB = 27 . ± .
06 mag and QT RGB = 27 . ± .
07 mag, where QT is a color-correctedF814W magnitude. From this we derive a distance modulus, ( m − M ) = 31 . ± . ± . d = 16 . ± . ± . Subject headings: galaxies: distances and redshifts — galaxies: stellar content — stars : Population II— galaxies: clusters: individual (Virgo, M60) — galaxies: elliptical and lenticular,cD INTRODUCTIONTwo of the popular distance indicators for nearby ellip-tical galaxies and early-type spiral galaxies are the sur-face brightness fluctuation (SBF) (Tonry & Schneider1988; Tonry et al. 2001; Blakeslee et al. 2009, 2010;Cantiello et al. 2011; Blakeslee 2012) and the plan-etary nebula luminosity function (PNLF) (Jacoby1989; Jacoby et al. 1990; Feldmeier et al. 2007;Teodorescu et al. 2011; Ciardullo 2012, 2013). TheSBF method is based on the fact that the variance in theimages of a galaxy depends on the distance to the galaxy.It can be applied to more distant galaxies compared withthe tip of the red giant branch (TRGB) method, butits precision decreases for the galaxies with compositestellar populations (Blakeslee 2012). The PNLF methodis based on the estimation of the [OIII] λ [email protected], [email protected] that the main calibrators for the SBF and PNLF areintermediate-type spiral galaxies, while the targets aremostly early-type galaxies. It is also noted that the dis-tance offset (between the SBF and PNLF distances) ver-sus distance modulus diagram (see Fig. 7 in Ciardullo(2012)) shows a trend that the scatter of this differencebecomes larger at ( m − M ) > .
0. To help resolve thisdiscrepancy, distance estimation based on another inde-pendent method is needed for the common targets of theSBF and PNLF methods.In this study we selected M60 (NGC 4649, VCC 1978),a giant elliptical galaxy in Virgo, to resolve the distancediscrepancy between PNLF distances and SBF distances.Basic parameters of M60 are listed in
Table 1 . M60has been a target of numerous studies because it showsseveral interesting features. First, it is the third bright-est elliptical galaxy in Virgo, and hosts a rich popula-tion of globular clusters, PNe, and low-mass X-ray bi-naries (LMXBs) (Lee et al. 2008a,b; Hwang et al. 2008;Strader et al. 2012; Pota et al. 2015; Teodorescu et al.2011; Luo et al. 2013; Mineo et al. 2014). Second, it hasa small spiral companion, NGC 4647 (SAB(rs)c), locatedat 2 . ′ M SMBH = 2 . × M ⊙ was found in this UCD, and provides a strongevidence that this UCD is not a globular cluster, but astripped nucleus of a genuine galaxy (Seth et al. 2014).Fourth, in the central region of M60 (51 . ′′ . ′′ TABLE 1Basic Parameters of M60
Parameter Value ReferencesR.A.(J2000), Dec(J2000) 12 h m s .0, +11 ◦ ′ ′′ B T = 10 . ± .
03 2Apparent total color ( B T − V T ) = 0 . ± .
01 2Ellipticity 0.11 3Position angle 71 deg 3 D ( B ) 262 ′′ . ′′ − A B = 0 . A V = 0 . A I = 0 .
040 5Distance modulus ( m − M ) = 31 . ± . ± . d = 16 . ± . ± .
42 Mpc 6Plate scale 78.7 pc arcsec − M TB = − . M TV = − .
85 2, 6Central velocity dispersion ( R eff /
8) 213 km s − R < R eff M = 1 . ± . × M ⊙ References . — (1) NED; (2) de Vaucouleurs (1991); (3) Makarov et al. (2014); (4)Cappellari et al. (2013); (5) Schlafly & Finkbeiner (2011); (6) This study; (7) Alabi et al.(2016). south), an underluminous Type Ia supernova, SN2004W,was discovered (Moore et al. 2004; Nielsen et al. 2012).Unfortunately the full light curve of SN2004W is notavailable. Fifth, it is a main member of the M60 group,which includes M60, NGC 4647, M59 (NGC 4621), NGC4660, and NGC 4638. This group may be the nearestcompact group of galaxies, if its nature as a genuinegroup is confirmed (Mamon 1989, 2008).In an extensive study of the PNe in M60,Teodorescu et al. (2011) determined a distance to thisgalaxy using a large sample of PNe, and presented( m − M ) = 30 . ± . . ± . m − M ) = 31 . ± . . ± . h m s m s s α (2000)11 ° ′ δ ( ) M60 NGC 4647R = 8 ’ (8 R eff ) Fig. 1.—
Location of the HST field (a large box) at R ≈ ′ inthe north from the center of M60 on the color image in the SDSS.North is up, and east to the left. A small square in the HST fieldindicates the location of a 10 ′′ × ′′ gray scale map shown in Figure2. A spiral galaxy in the north-west of M60 is NGC 4647. WhetherNGC 4647 is interacting with M60 has been controversial. the final section, main results are summarized. DATA AND DATA REDUCTION
Figure 1 displays a finding chart for a 20 ′ × ′ fieldincluding M60 based on the SDSS color map. It showsalso NGC 4647, a spiral galaxy in the northwest of M60.We used the ACS/WFC F775W (SDSS i ′ ) and F850LP(SDSS z ′ ) images of an outer field in the north of M60,the location of which is marked in Figure 1 . Theseimages were obtained as part of the ACS Pure ParallelProgram (PID:9575, PI. William Sparks) in 2002. How-ever, they are very useful for the study of the resolvedRGB Distance to M60 3 ’’
236 pc
M60, F775W
Fig. 2.—
A grayscale map of F775W image of a 10 ′′ × ′′ fieldfor M60, which is marked by a small square in Figure 1. Most ofthe point sources in this map are old red giant stars belonging toM60. stars in M60 as well.Since the difference between the effective wavelengthsof F775W and F850LP filters is small, the combinationof these two filters is not effective for the study of col-ors (or metallicity) of stellar populations. However thiscombination of the filters is good enough to detect redgiants in nearby galaxies.The HST field is located at ≈ ′ in the north from M60in the sky. The effective radius of M60 is R eff = 58 . ′′ R eff . Therefore thecrowding of the point sources in this field is much lowercompared with inner fields so that this field is much moresuitable for the study of resolved stars in M60. We com-bined individual exposure images to produce deep masterimages using the AstroDrizzle package. Total exposuretimes are 15,407 s for F775W and 9,547 s for F850LP sothat the images are deep enough to study the resolvedstars in M60. A gray scale map of the F775W image fora 10 ′′ × ′′ section of the entire field (marked by a smallsquare in Figure 1 ) is shown in
Figure 2 . In the figure,many point sources are clearly seen. Most of them arered giant stars belonging to M60, and some of them maybe compact background galaxies.We obtained photometry of the point sources in theimages using the latest version of DOLPHOT (Dolphin2000). We used charge transfer efficiency corrected andflat-fielded images (* flc.fits images) with the syntheticTiny Tim point spread functions (PSFs) (Krist et al.2011). The DOLPHOT parameters used in this studyare the same as those given in DOLPHOT/ACS user’sguide (version 2.0).We carried out artificial star tests using the artifi-cial star routine (acsfakelist) in DOLPHOT. We gen-erated a sample of artificial stars having a color rangeof F775W– F850LP = 0.3 ∼ ∼ R ec ov e r y r a t e [ % ] -0.20.00.2 ∆ F L P
24 25 26 27 28F850LP
Input -0.20.00.2 ∆ C o l o r (a)(b)(c) Fig. 3.— (a) Recovery rates of red giant branch stars with 0 . ≤ (F775W–F850LP) ≤ .
9, derived from artificial star experiments.(b) Differences between input and output F850LP magnitudes (in-put minus output) as a function of input F850LP magnitudes. Me-dian offsets and standard deviations in each magnitude bin are in-dicated by red dots and vertical lines, respectively. (c) Same as (b)except for (F775W–F850LP) colors. artificial stars, which corresponds to ∼
10% of the to-tal number of detected sources, in each image, and car-ried out PSF photometry as done on the original frames.We iterated this procedure 50 times to reduce statisti-cal uncertainties.
Figure 3 displays the recovery ratesof the input stars, and the difference in F850LP mag-nitudes and (F775W-F850LP) colors between the in-put and output values. It shows that 50% complete-ness limit is F850LP ∼ − . ± .
007 mag and ∆(F775W–F850LP) = 0 . ± .
007 mag, for F850LP = 26.8 magwhich is close to the TRGB magnitude.Since the TRGB calibration is based on
V I (orF606W and F814W in the HST system) (Lee et al. 1993;Rizzi et al. 2007; Jang & Lee 2017a), we need to trans-form F775W and F850LP photometry to F606W andF814W photometry. For this purpose we used the 12Gyr isochrones in the Dartmouth model (Dotter et al.2008). In
Figure 4 we plotted the color-color re-lations for the TRGB of the isochrones for a rangeof metallicity ( − . ≤ [Fe/H] ≤ . . < .
35. From the linear fits for thedata, we obtain Lee & Jang 2017
TRGB (Dartmouth) F W - F W F W - F L P (a)(b) [ F e / H ] = - . -0.2-1.0 Lupton+05
Fig. 4.— (a) (F606W–F814W) vs. (F775W–F850LP) rela-tion. Pentagons denote the TRGB color from the 12 Gyr Dart-mouth model with − . ≤ [Fe/H] ≤ . ( F W − F W ) = (3 . ± . F W − F LP ) − (0 . ± . rms = 0 .
034 for (F775W–F850LP) ≤ .
9, and( F W − F W ) = (1 . ± . F W − F LP )+(1 . ± . rms = 0 .
029 for (F775W–F850LP) > . F W − LP ) = (0 . ± . F W − F LP )+(0 . ± . rms = 0 . I = i − . × ( i − z ) − . Fig-ure 4(b) . The second relation derived in this study isvery similar to this transformation, except for the slightoffset in the blue end. Using the equations above, we cantransform (F775W–F850LP) colors and F850LP magni-tudes of the detected stars in the HST field of M60 to(F606W–F814W) colors and F814W magnitudes. RESULTS3.1.
CMDs of the Resolved Stars in M60 In Figure 5 we plotted the F850LP–(F775W–F850LP) CMD of the detected point sources in the HSTfield of M60. The most prominent feature in the CMD is
M60, Entire field0.0 0.5 1.0 1.5 2.0F775W-F850LP28272625 F L P Fig. 5.—
F850LP–(F775W-F850LP) color-magnitude diagramfor the resolved point sources in the HST field of M60. Curvedlines represent 12 Gyr stellar isochrones with a range of metallicity([Fe/H] =–2.2 to 0.0, in steps of 0.2, from left to right) in the Dart-mouth models (Dotter et al. 2008). The vertical shaded region at0.3 ≤ (F775W–F850LP) ≤ a concentration of red stars with a broad range of colors,the mean value of which is (F775W–F850LP) ≈ .
6. Itis a red giant branch (RGB) of M60. The brightest partof this RGB is seen at F850LP ≈ . < . TRGB Distance Estimation
We determine a TRGB distance to M60 from photom-etry of the resolved stars, as done in our previous studiesfor other galaxies (Lee & Jang 2016; Jang & Lee 2017b).
Table 2 lists a summary of TRGB distance estimationfor M60. First, we selected the blue red giant candidatesinside the shaded region in the CMD of
Figure 5 . TheTRGB magnitude is almost constant in this blue RGB.The data for M60 used in this study are not deep enoughto cover the full range of colors of the RGB stars. In thiscase, using the blue RGB is the best way to avoid anycomplications due to redder stars. Then we derived theirluminosity functions as shown in
Figure 6 . Applying theRGB Distance to M60 5 N F L P T R G B = . ± . M60
Fig. 6.—
F850LP luminosity function for blue RGB stars in M60(histogram) and corresponding edge detection response (red line).The position of the TRGB is marked by the dashed line. edge-detection method with a Sobel kernel [–1, –2, –1, 0,1, 2, 1] to this luminosity function, we calculated theedge-detection responses, as plotted by the red solid linein the figure. The edge-detection response shows clearlya major single peak at F850LP ∼ T RGB = 26 . ± .
06 mag. The median color of the TRGB,(F775W–F850LP)
T RGB = 0 . ± .
02, is measured usingthe RGB stars at ± α = 0 . T RGB = 26 .
70 mag and (F775W–F850LP)
T RGB = 0 .
60. We derived the TRGB magnitude and color fromthe recovered artificial stars using the same procedure.The mean values of the input minus output TRGB mag-nitudes and colors are ∆F850LP
T RGB = − . ± . T RGB = 0 . ± .
02 mag.Correcting the measured TRGB values with these sys-tematic offsets, we obtain F850LP
T RGB = 26 . ± . T RGB = 0 . ± . A F775W = 0 .
043 and A F850LP = 0 . TABLE 2A Summary of TRGB Distance Measurement for M60
Parameter ValueApparent TRGB magnitude in F850LP 26 . ± . . ± . − . ± . . ± . . ± . . ± . . ± . . ± . . ± . . ± . QT . ± . − . ± . m − M ) . ± . r ± . s Distance, d [Mpc] 16 . ± . r ± . s ternal reddening for M60 in this analysis. We convertedthe measured TRGB magnitude and color in the F775Wand F850LP system to the F606W and F814W systemusing the photometric transformations described in Sec-tion 2, obtaining F814W ,T RGB = 27 . ± .
06 mag and(F606W–F814W) ,T RGB = 2 . ± . I -band TRGB has a weakmetallicity dependence, especially at the red colorrange (F606W–F814W & QT magnitude. It is describedby QT = F814W − . Color − . + 0 . Color − . Color = (F606W–F814W) . The absolutezero-point of the QT is measured to be M QT,T RGB = − . ± .
056 mag, from the combination of two distanceanchors with known geometric distances (NGC 4258 andthe LMC). The systematic error of ± .
056 in this cali-bration is much smaller than the values given in the pre-vious studies. The value of the QT magnitude and cor-responding distance modulus for M60 we obtained are: QT RGB = 27 . ± .
07 mag and ( m − M ) = 31 . ± . . ± .
50 Mpc). The systematicuncertainty of this distance modulus is ± .
056 mag (cor-responding to the distance error of ± .
42 Mpc). DISCUSSION4.1.
Comparison of the TRGB distance to M60 withPNLF and SBF Distances
We compared our TRGB distance estimate for M60with those based on the PNLF and SBF in the literature,as summarized in
Table 3 and plotted in
Figure 7 .Jacoby et al. (1990) presented a PNLF distance to M60derived from a small sample of 16 PNe, ( m − M ) =30 . ± .
14 ( d = 14 . ± . m − M ) =30 . +0 . − . by Ciardullo et al. (2002). They adopted thecalibration for the PNLF given by Ciardullo et al. (1989), M ∗ PN = − .
48, which is based on the Cepheid distance toM31, 710 kpc (( m − M ) = 24 . ± .
10) and foregroundreddening E ( B − V ) = 0 . ± .
02. The distance to M31adopted for this PNLF calibration is somewhat smallerthan the values in more recent Cepheid distances to M31: Lee & Jang 2017
TABLE 3Comparison of Distance Estimates for M60
Method ( m − M ) References RemarksTRGB 31 . ± . ± . . ± . ± . M ∗ PN = − . . +0 . − . Ciardullo et al. (2002) N(PN)=16, M ∗ PN = − . . ± . M ∗ PN = − . . ± . M ∗ PN = − . ± . . ± .
09 MeanSBF 31 . ± .
11 Neilsen & Tsvetanov (2000) F814W31 . ± .
15 Tonry et al. (2001) I . ± . ± . . ± . ± . . ± .
05 Mean P N L F m ea n S B F m ea n Jacoby+90Ciardullo+02Teodorescu+11Ciardullo+13 This study Neilsen+00Tonry+01Mei+07Blakeslee+09
Fig. 7.—
Comparison of the TRGB distance (red pentagon) inthis study with the PNLF (blue pentagons) and SBF distances(yellow pentagons) in the literature. Error bars indicate standarderrors. Mean values of the PNLF and SBF distances are markedby the blue and yellow vertical strips, respectively. ( m − M ) = 24 . ± .
08 and ( m − M ) = 24 . ± . m − M ) = 30 . ± .
20, including a systematic er-ror of 0.13. They adopted the same calibration for thePNLF as used in Jacoby et al. (1990). If the metal-licity dependence of the period-luminosity relation forCepheids is adopted for bright galaxies, this calibra-tion will be slightly brighter to M ∗ PN = − . ± . M ∗ PN = − . ± . m − M ) = 30 . ± . m − M ) = 31 . ± .
11. This value is consistent with the value given by Tonry et al. (2001), ( m − M ) = 31 . ± .
05. LaterMei et al. (2007) presented a similar value, ( m − M ) =31 . ± .
07. Blakeslee et al. (2009) updated the dis-tances in Mei et al. (2007) with an improved calibra-tion, presenting a 0.11 mag smaller value for M60,( m − M ) = 31 . ± .
08. This is the most recent valuefor the SBF distance to M60 available in the literature.The mean value of these SBF distances is derived to be( m − M ) = 31 . ± .
05. Our TRGB distance modulus,( m − M ) = 31 . ± .
09, is ∼ ∼ Interaction between M60 and NGC 4647
NGC 4647, a spiral galaxy, is located only 2 . ′ ± − , NED) is only about 300 km s − larger thanthat of M60 (1110 ± − , NED). Because of theprojected proximity and the small radial velocity differ-ence of M60 and NGC 4647, several studies investigatedany possibility of tidal interaction between these twogalaxies (de Grijs & Robertson 2006; Lanz et al. 2013;D’Abrusco et al. 2014; Mineo et al. 2014; Pota et al.2015). However whether NGC 4647 is interacting withRGB Distance to M60 7M60 or not is still controversial (de Grijs & Robertson(2006), Pota et al. (2015) and references therein).Optical images of this pair of galaxies, in Figure 1 ,show little evidence for any significantly distorted struc-tures around each galaxy, although they are close to eachother in the sky. This indicates two possibilities. First,the relative distance along the line of sight in the spacebetween the two galaxies is so large that they are not in-teracting. Second, they are relatively close to each otherin the space, but their interaction is weak. RecentlyPota et al. (2015) found, from the study of kinematicsof the globular clusters in M60, no strong evidence tosupport the interaction between M60 and NGC 4647.They suggested that M60 and NGC 4647 may be only inthe beginning stage of interaction, if they are interacting,as suggested earlier by de Grijs & Robertson (2006) whonoted a presence of weak young blue stellar population inthe north-west direction of M60 in the HST/ACS imagesof the central region.It is known that strongly-interacting galaxies show rel-atively stronger MIR and FIR emission than weakly-interacting galaxies so the spectral energy distributions(SEDs) of galaxies are useful to estimate the stage ofinteraction (Lanz et al. 2013). Dopita et al. (2002) pre-sented a five-stage classification scheme to estimate thestage of galaxy interaction. According to this scheme,weakly-interacting galaxies (Stage 2) show minimal mor-phological distortions, and moderately-interacting galax-ies (Stage 3) show stronger morphological distortion in-cluding often tidal tails (Lanz et al. 2013). Lanz et al.(2013) suggested, from the SEDs of NGC 4647 and M60based on UV to FIR data, that this pair of galaxies is inthe moderately interacting stage.We need to know a relative distance between M60 andNGC 4647 to conclude on the possibility of tidal inter-action between the two galaxies. Unfortunately there isnot yet any TRGB distance to NGC 4647. There areseveral estimates for the distance to NGC 4647 basedon the Tully-Fisher relation in the literature, showing alarge range: ( m − M ) = 31 . ± .
54 to 31 . ± . m − M ) = 31 . ± .
26 ( d = 17 . ± . m − M ) = 31 . ± . d = 16 . ± . ± . SUMMARYWe present photometry of the resolved stars in an outerfield of M60 based on deep F775W and F850LP images.This is the first photometry of the resolved giant stars inM60. Primary results in this study are summarized asfollows.1. The CMD of the resolved stars in M60 shows adistinguishable broad RGB. A TRGB is detected atF850LP
TRGB = 26 . ± .
06 mag, in the luminosityfunction of the red giants. This value correspondsto F814W , TRGB = 27 . ± .
06 mag and QT RGB =27 . ± .
07 mag.2. From the magnitude of the TRGB we derive a dis-tance modulus, ( m − M ) = 31 . ± . ± . ± . d = 16 . ± . ± . ± .
65 Mpc).3. The TRGB distance modulus for M60 derived inthis study is 0.3 mag larger than the mean PNLFdistance values, and 0.1 mag smaller than the SBFdistance values. This indicates that the PNLF dis-tances to M60 in the literature have larger uncer-tainties than the suggested values.4. We checked the relative distance between M60 andNGC 4647, a nearby spiral galaxy, to investigateany tidal interaction between the two galaxies. Itis found that the TRGB distance to M60 andthe Tully-Fisher distance to NGC 4647 are simi-lar within the errors. However, absence of any sig-nificantly distorted structures around each galaxyindicates that the relative distance between the twois not close enough to show strong tidal interaction.This study is supported by the National ResearchFoundation of Korea (NRF) grant funded by the KoreaGovernment (MSIP) (No. 2012R1A4A1028713). Thispaper is based on image data obtained from the Multi-mission Archive at the Space Telescope Science Institute(MAST).
REFERENCESAlabi, A. B., Forbes, D. A., Romanowsky, A. J., et al. 2016,MNRAS, 460, 3838Bellazzini, M., Ferraro, F. R., & Pancino, E. 2001, ApJ, 556, 635Bird, S., Harris, W. E., Blakeslee, J. P., & Flynn, C. 2010, A&A,524, A71Blakeslee, J. P., Jord´an, A., Mei, S., et al. 2009, ApJ, 694, 556Blakeslee, J. P., Cantiello, M., Mei, S., et al. 2010, ApJ, 724, 657Blakeslee, J. P. 2012, Ap&SS, 341, 179Caldwell, N. 2006, ApJ, 651, 822Cantiello, M., Biscardi, I., Brocato, E., & Raimondo, G. 2011,A&A, 532, A154 Cantiello, M., Grado, A., Blakeslee, J. P., et al. 2013, A&A, 552,A106Cappellari, M., McDermid, R. M., Alatalo, K., et al. 2013,MNRAS, 432, 1862Ciardullo, R., Jacoby, G. H., Ford, H. C., & Neill, J. D. 1989,ApJ, 339, 53Ciardullo, R., Feldmeier, J. J., Jacoby, G. H., et al. 2002, ApJ,577, 31Ciardullo, R. 2012, Ap&SS, 341, 151Ciardullo, R. 2013, Advancing the Physics of Cosmic Distances,289, 247D’Abrusco, R., Fabbiano, G., Mineo, S., et al. 2014, ApJ, 783, 18