1: %\documentclass[preprint2]{aastex} Nov 28, June 13, 2007

2: \documentclass[12pt,preprint]{aastex}

3: %\usepackage{apjfonts}

4: 

5: \def\as{{$''\!\!$.}\/}

6: \def\mp{{$^{\rm m}\!\!$.}\/}

7: \def\lea{\mathrel{<\kern-1.0em\lower0.9ex\hbox{$\sim$}}}

8: \def\gea{\mathrel{>\kern-1.0em\lower0.9ex\hbox{$\sim$}}}

9: 

10: %\slugcomment{Submitted to the ApJL on \today }

11: \slugcomment{Accepted for Publication in AJ}

12: %\slugcomment{Updated on \today}

13: \shorttitle{Star Clusters in M51} \shortauthors{Hwang \& Lee}

14: 

15: \begin{document}

16: 

17: \title{A Catalog of Bright Star Clusters in the Interacting Galaxy M51\altaffilmark{*}}

18: 

19: \author{\sc Narae \ Hwang and Myung Gyoon \ Lee}

20: \affil{Astronomy Program, Department of Physics and Astronomy,\\

21: Seoul National University, Seoul 151-747, Korea}

22: \email{nhwang@astro.snu.ac.kr, mglee@astrog.snu.ac.kr}

23: 

24: \altaffiltext{*}{Based on observations made with the NASA/ESA {\it

25: Hubble Space Telescope}, obtained from the data archive at the

26: Space Telescope Science Institute, which is operated by the

27: Association of Universities for Research in Astronomy, Inc., under

28: NASA contract NAS 5-26555.}

29: 

30: 

31: \begin{abstract}

32: We present a catalog of star clusters with

33: $V_{F555W}<23$ mag detected in the interacting spiral galaxy M51

34: system based on the mosaic images taken with $HST$ ACS by the

35: Hubble Heritage Team. We have selected about 3,600 clusters based

36: on their morphological information through the visual inspection.

37: The final star cluster catalog

38: includes 2,224 clusters that are relatively well isolated and have

39: a circular shape. The star clusters in M51 are mostly distributed

40: around the spiral arms of NGC 5194. The color-magnitude diagrams

41: show that most of the star clusters in M51 are bluer than

42: $(B_{F435W}-V_{F555W}) = 0.5$ and $(V_{F555W}-I_{F814W}) = 0.8$.

43: There are also some red star clusters with $(B_{F435W}-V_{F555W})

44: > 0.7$, uniformly distributed over the M51 field. Some of these red

45: clusters are suspected to be a part of the halo or old disk population

46: based on their old ages ($t \geq 10^9$ yrs) and their spatial

47: distribution. The luminosity function of the star clusters is

48: fit well by a single power law 

49: with $\alpha = -2.59 \pm 0.03$ for the range $-10.0 <M_V < -8.0$ mag.

50: We find that the size distribution of the star clusters can be fit with

51: three Gaussian components with peaks at effective radii of 2.27, 4.80 and 7.51 pc.

52: Some large star clusters with red color are faint

53: fuzzy clusters, and they are distributed not only around NGC 5195

54: but also around NGC 5194. These faint fuzzies are found to display

55: an elongated spatial distribution, while the normal compact red

56: clusters show a relatively uniform distribution around NGC 5194.

57: 

58: 

59: \end{abstract}

60: 

61: \keywords{galaxies: individual (M51; NGC5194; NGC 5195) ---

62: galaxies: spiral

63: --- galaxies: interaction --- galaxies: evolution --- galaxies:

64: star clusters }

65: 

66: 

67: \section{Introduction}

68: 

69: 

70: Studying star clusters is important for understanding the star

71: formation history of their host galaxies and the star cluster population

72: itself. Globular clusters found in giant elliptical galaxies show

73: that most elliptical galaxies are very old systems that ceased to

74: form stars or star clusters a long time ago (see Lee 2003 and

75: references therein). On the contrary, many star clusters in

76: late-type galaxies display rather blue colors, indicating very

77: recent star formation activities in the corresponding galaxies

78: \citep{lee06}. For interacting or merging galaxies, it is known

79: that there are some very bright and massive star clusters such as

80: super-star clusters (SSCs) in Antennae galaxy \citep{whi99}. These

81: massive clusters are found in galaxies that are undergoing an

82: active star formation probably induced by dynamical interactions.

83: 

84: The existence of such massive star clusters indicates that burst-like star formation

85: activities in some interacting or merging galaxies are intimately

86: correlated with the dynamical events undergone by their host

87: galaxies. This also suggests that, with regard to normal galaxies,

88: star clusters can also be used as an excellent tracer of star

89: formation activity in interacting or merging galaxies. However,

90: finding star clusters in galaxies is not easy. This is possible

91: only for nearby galaxies even after the advent of the {\it Hubble

92: Space Telescope} (HST), which provides a superb spatial resolution

93: and photometric depth. M51 is located at a distance of 8.4 Mpc

94: \citep{fel97} so that it is one of the most suitable

95: interacting galaxies for star cluster survey using the $HST$.

96: 

97: M51 is an interacting galaxy system that is composed of a face-on

98: Sbc-type spiral galaxy NGC 5194 and a fainter companion, the SB0

99: galaxy NGC 5195. It is abundant with interstellar media

100: (see a recent study by \citet{schu07}). Several conditions

101: with M51 make it suitable for the star cluster study using the $

102: HST$. (1) It is possible to marginally resolve the star clusters

103: at the distance of M51 with the resolving power of the $HST$. (2)

104: M51 is in a nearly face-on configuration, which allows a direct

105: view of the entire disk structure. (3) There are some previous

106: theoretical studies \citep{too72,sal00a,sal00b} suggesting that

107: these two galaxies may have experienced a single or multiple

108: encounters some hundreds of Myrs ago. That is, the properties and

109: epoch of the dynamical events are known to some extent, which

110: provides the base timeline for comparison with the derived star

111: formation history.

112: 

113: There are several previous studies on the star cluster population

114: in M51 that have used $HST$ data. In \citet{bik03} and

115: \citet{bas05a}, approximately 1,000 star clusters were detected in

116: a small region near the M51 center based on $HST$ WFPC2 data.

117: \citet{bas05a} argued that the cluster formation rate seems to

118: increase considerably around the expected time of the second

119: encounter according to the derived age distribution of the star

120: clusters. It was also suggested that young and massive star clusters in

121: M51 tend to form in complexes rather than in isolation

122: \citep{bas05b} and even clusters as massive as $10^6 M_{\odot}$

123: can be disrupted in the M51 disk within a few Gyrs after their

124: formation \citep{gie05}.

125: 

126: On the other hand, \citet{lee05} adopted a conservative policy for

127: star cluster selection and reported only about 400 resolved star clusters

128: based on $HST$ WFPC2 data that covered the larger area than used in the earlier studies.

129: They showed that the

130: age distribution of these resolved clusters displays a broad peak

131: at several hundred Myrs, and this is consistent with the epoch of

132: the first encounter predicted by the theoretical models. Another

133: important point noted in \citet{lee05} is that the star cluster

134: sample can be so easily contaminated that star clusters should be

135: selected with great care for the case of M51, since they are only

136: marginally resolved and can be easily confused with other sources

137: such as bright point sources.

138: 

139: Previous studies on the star clusters in M51 have a common limit

140: regarding the $HST$ data used, which is the incomplete areal

141: coverage of M51. Since the existing $HST$ WFPC2 observation of M51

142: only covered the central parts of NGC 5194 and NGC 5195 along with

143: some parts of the spiral arms, the star cluster sample reported in

144: the previous studies can be biased toward young and massive

145: clusters mostly found around the galaxy center and the spiral

146: arms. However, in order to investigate the complete star formation

147: history using the star clusters in M51, we need to carry out a

148: star cluster survey over the field of entire galaxy.

149: Although the globular cluster search in M51 performed by

150: \citet{cha04} returned 34 globular cluster candidates, it still

151: suffers from an incomplete areal coverage.

152: 

153: We have carried out a star cluster survey over the

154: M51 field using the newly released Hubble Heritage Data that

155: covers the main body of NGC 5194.

156: As a first step to the study of the star cluster

157: population in M51, we have searched for both old (red) and young

158: (blue) star cluster candidates based on a homogeneous detection

159: and classification policy that employs the visual inspection of

160: sources before the final registration of the star cluster

161: candidates. This method does not guarantee the complete census of

162: the star clusters due to the inherent limitation of manual

163: inspection. However, this approach can be more efficient to lower

164: the possible contamination in star cluster samples, as shown by

165: \citet{lee05}. Some preliminary results of this study showed the

166: overabundance of faint fuzzy clusters around NGC 5195

167: \citep{hwa06} and implied a possible correlation between the

168: dynamical encounters of host galaxies and age distribution of star

169: cluster candidates \citep{hwa07a}.

170: 

171: In this paper, we present the results of M51 star cluster survey,

172: including the methodology of finding and classifying star cluster

173: candidates. We also present the resulting final catalog of star

174: cluster candidates with $V_{F555W}<23$ mag. This cluster catalog

175: has served as the master catalog in \citet{hwa06,hwa07a}. Then, we

176: investigate some photometric properties and the spatial

177: distribution of these cluster candidates. Finally, some properties

178: of large clusters such as faint fuzzy clusters found in the M51

179: system are discussed. During this study, there came a

180: report by \citet{sch07} that identified about 1300

181: star cluster candidates in M51 based on the same Hubble Heritage

182: Data. We present an independent and complementary study that was

183: carried out by using a different approach. We will discuss some

184: common characteristics of star clusters shown in both studies at

185: the end of this paper.

186: 

187: 

188: \section{Observation and Data}

189: 

190: The data used in this study were obtained by the Hubble Heritage

191: Team to commemorate the $HST$'s 15th anniversary using the $HST$

192: ACS with F435W, F555W, F814W, and F658N filters through $HST$

193: observing program 10452 (P.I.: Steven V. W. Beckwith). The

194: observed field is as large as about $6.8' \times 10.5'$ centered

195: on NGC 5194. The accumulated exposure times are 2,720 s for the

196: F435W band and 1,360 s for the F555W and F814W bands. All the

197: necessary data reduction processes were done by the STScI

198: including multi-drizzling and image combination before the data

199: release. More detailed information on the observation and data

200: reduction is given in \citet{mut05}. \footnote{See also

201: http://archive.stsci.edu/prepds/m51/.}

202: 

203: We adopted a distance of $8.4\pm0.6$ Mpc ($(m-M)_0=29.62$) to M51

204: determined from the planetary nebula luminosity function in M51

205: \citep{fel97}. The corresponding linear scale is 40.7 parsecs per

206: arcsecond. Since the $HST$ ACS mosaic data have pixel sizes of

207: $0.05\arcsec$, the resultant projected physical scale covered by

208: one pixel is approximately $2.04$ pc. The foreground reddening

209: toward M51 is low, $E(B-V)=0.035$, and the corresponding

210: extinctions are $A_B=0.150$, $A_V=0.115$, and $A_I=0.067$ mag

211: \citep{sch98}. The total magnitudes and colors are

212: $B^T=8.96\pm0.06$ mag and $(B^T-V^T)=0.60\pm0.01$ for NGC 5194,

213: and $B^T=10.45\pm0.07$ mag and $(B^T-V^T)=0.90\pm0.01$ for NGC

214: 5195 \citep{dev91}. At the adopted distance (without internal

215: extinction correction), the absolute magnitudes are $M_B^T=-20.81$

216: mag, $M_V^T=-21.38$ mag for NGC 5194, and $M_B^T=-19.32$ mag and

217: $M_V^T=-20.19$ mag for NGC 5195.

218: 

219: 

220: \section{Star Cluster Detection and Classification}

221: 

222: \subsection{Source Detection and Photometry}

223: \label{phot}

224: 

225: Detecting a source is a demanding task if the source is

226: embedded in a varying background, such as the spiral arms of M51.

227: To work around this difficulty, a representative background

228: map was constructed by median filtering of the F555W band image of

229: M51. The $RMEDIAN$ task of $IRAF$ package was employed for this

230: process by setting the inner and outer radii of the filtering ring

231: to 20 and 30 pixels, respectively. Then, the constructed

232: background map was subtracted from the original F555W band image

233: of M51 to produce a detection image in which only the resolved

234: sources remain. Source detection was done by running SExtractor

235: \citep{ber96} on this detection image. A detection threshold of

236: $4\sigma$ and minimum contiguous detected area of five pixels were

237: used for finding sources in the detection image. The total number

238: of detected sources is approximately 120,000.

239: 

240: The flux of the detected objects was measured using SExtractor run

241: in the dual mode in the original F435W, F555W, and F814W band

242: images. Among the various photometric measurement values produced

243: by SExtractor, we adopted a flux measured using an aperture with

244: $r = 6$ pixels ($0.3\arcsec \simeq 12$ pc), which is about three times the

245: average FWHM of the point sources in the images, as the

246: representative flux of an object. The colors of the objects were

247: also calculated based on the photometric magnitudes measured with

248: the same $r=6$ pixel apertures in the corresponding bands. The

249: instrumental magnitudes in the F435W, F555W, and F814W bands were

250: calibrated using the photometric zero points of the Vega magnitude

251: system for the WFC of $HST$ ACS provided by \citet{sir05}.

252: Hereafter, we use $B$, $V$, and $I$ to denote $B_{F435W}$,

253: $V_{F555W}$, and $I_{F814W}$, respectively.

254: 

255: 

256: \subsection{Photometry Comparison with Previous Studies}

257: 

258: There are two sets of photometry data for the M51 star cluster

259: candidates available in the literature: one from \citet{bik03} and the other from

260: \citet{lee05}. Since \citet{lee05} have already shown that their

261: photometry measurements are in good agreement with those in

262: \citet{bik03}, we used the more recent photometry data from

263: \citet{lee05} for comparison purposes. In \citet{lee05}, aperture

264: photometry was derived using the aperture of $r = 3$ pixels and

265: was corrected for the finite aperture effect by adding $-0.39$ mag for

266: PC and $-0.31$ mag for WF. The aperture size of $r=3$ pixels in WF

267: of WFPC2 corresponds to 6 pixels ($0.3 \arcsec$) in ACS, which is

268: our adopted aperture size for photometry. We compared our

269: photometry data with that of WF CCDs in \citet{lee05} after

270: removing the aperture corrections.

271: 

272: Figure \ref{leecomp} shows the result of the photometry

273: comparison. Our photometric measurements are in reasonably good

274: agreement with those of \citet{lee05}. The mean differences in

275: photometry for about 230 common objects are $\Delta V = -0.039 \pm

276: 0.063$ mag, $\Delta (B-V) = 0.048 \pm 0.055$ mag, and $\Delta

277: (V-I) = -0.040 \pm 0.057$ mag, where $\Delta$ means \citet{lee05}

278: minus this study, and the errors are standard errors of the mean.

279: 

280: \subsection{Star Cluster Classification}

281: \label{visual}

282: 

283: At the distance of M51, typical star clusters appear to be

284: somewhat more extended than stars in the $HST$ images (see Figure

285: 2 of \citet{lee05}). Therefore, it is possible to select cluster

286: candidates using the morphological parameters provided by

287: SExtractor. We used the cluster sample of \citet{lee05} as the

288: training set for constraining the parameter spaces of the objects'

289: FWHM, ellipticity (hereafter `$e$'), and stellarity in which most

290: of the star clusters would be found. This test revealed that most

291: of the star clusters in \citet{lee05} are found to belong to two

292: groups in the parameter spaces of FWHM, $e$, and stellarity:

293: Group 1 with low stellarity $< 0.5$, $2.4 < FWHM < 20$ pixels, and $e < 0.6$,

294: and Group 2 with high stellarity $> 0.9$, $2.4 < FWHM < 40 $ pixels,

295: and $e < 0.6$. The number of selected candidates

296: with $V < 23$ mag is 7,300 (about 5,900 for Group 1 and about

297: 1,400 for Group 2). For a comparative reference, about 1,100 point

298: sources with $V < 23$ mag, $1.0<FWHM<2.3$ pixels, $e<0.2$, and

299: stellarity $>0.9$ are also selected. Therefore, about 8,400

300: out of the approximately 120,000 detected sources are

301: selected for the final visual classification.

302: 

303: The star cluster candidates in M51 can be found in various shapes

304: and environments. Therefore, during the visual classification, we

305: considered the following two factors: (1) the width and curvature

306: of the radial profiles and (2) the shape of the 2D contour and

307: environmental effects. The width and curvature of the radial

308: profiles are the most basic parameters used to select the star

309: cluster candidates. These criteria mainly serve to distinguish

310: star cluster candidates from point sources. The FWHM of the star

311: cluster candidates in M51 is about 3 pixels or larger, while the

312: FWHM of the point sources is about $2.1 - 2.3$ pixels in $HST$ ACS

313: data. The large FWHM of the star cluster candidates is also

314: reflected in the curvature of the radial profiles. The radial

315: profiles of the point sources show a very steep decrease away from

316: the core and they reach the background level within a radius of

317: 2.5 pixels. On the other hand, the radial profiles of the star

318: cluster candidates have a relatively gentle slope and extend to a

319: radius of more than 3 pixels.

320: 

321: The shape of the 2D contour reveals how circular a star cluster

322: candidate appears and how many subcomponents it has as.

323: We separate a sample of star cluster

324: candidates with circular shapes from another sample that shows

325: distorted contours and/or multiple peaks in a single contour.

326: Another important information that can be derived from the 2D

327: contour map is how many neighbors there are near the candidate.

328: That is, we can investigate if a star cluster candidate is

329: surrounded by several sources that are clustered together or

330: whether a candidate has distorted contours or many subcomponents,

331: indicating the existence of very nearby unresolved sources or

332: multiple sources that are merged together. This helps to separate

333: well isolated star clusters with circular shape from those with

334: elongated or irregular shape and/or any prominent neighbors.

335: 

336: >From the visual classification based on these criteria, we define

337: a `Class 1' cluster sample as `cluster candidates' that have a very

338: well-defined nearly circular shape, and no prominent nearby neighbors.

339: Another class of star cluster candidates defined is a `Class 2'

340: cluster sample that includes `cluster candidates' with an elongated shape as 

341: well as irregular structures and/or multiple neighbors.

342: Figure \ref{contour} shows 2D contour maps of a few Class 1 and Class 2

343: cluster samples. After the visual

344: inspection of about 8,400 candidate extended candidates and point

345: sources, we have selected 2,224 Class 1 and 1,388 Class 2 star

346: cluster candidates.\footnote{Hereafter, Class 1 and Class 2 star

347: cluster candidates will be referred to as just `star clusters'

348: even though they are not yet spectroscopically confirmed to be genuine

349: star clusters.} It turns out that about $50\%$ of Group 1

350: candidates are classified as Class 1 ($\sim 35\%$) or Class 2

351: ($\sim 15\%$) clusters, while about $23\%$ of Group 2 are

352: classified as Class 1 ($\sim 16\%$) or Class 2 ($\sim 7\%$)

353: clusters.

354: 

355: \subsection{Aperture Correction}

356: \label{apc}

357: 

358: The aperture photometry of sources requires the correction for the

359: light lost due to the finite size of the aperture used. For point

360: sources such as stars, the amount of aperture correction is

361: determined by the instrumental characteristics of $HST$.

362: Therefore, we can apply a single value of aperture correction to

363: the photometry data of point sources. However, for extended

364: sources such as star clusters in M51, a single value of correction

365: cannot be defined based on a certain size of the aperture since

366: the luminosity profile of one cluster varies depending on its size and compactness.

367: Figure \ref{apc1} shows the integrated magnitude profiles of 16

368: star clusters in M51, which are found in well-isolated locations

369: and have nearly circular shapes. The effective radii $R_{\rm eff}$ of

370: these clusters range from 2.85 pc to 8.19 pc with a median value

371: of 5.01 pc. It is evident that the shape of the profile depends on

372: the size and/or the compactness of a cluster, which varies from

373: one cluster to another. This leads to a different amount of

374: aperture correction for each star cluster.

375: 

376: We investigated the correlation between the slope of the

377: integrated magnitude profile around the fixed radius of $r = 6$

378: pixels, the same size as that of the aperture used for the

379: photometry, and the amount of necessary correction calculated to

380: the outer radius of $r = 18$ pixels for each star cluster. The

381: slope of the integrated magnitude profile for each cluster at

382: $r=6$ pixels was derived from the difference between the $V$-band

383: integrated magnitudes for $r=5$ and $7$ pixels. As shown in Figure

384: \ref{apc2}, there is a good correlation between the slope of the

385: profile and the amount of magnitude correction $\Delta Mag$. The

386: relationship between the slope and $\Delta Mag$ is derived by

387: linear fitting: $\Delta Mag = 4.72 \times slope - 0.01$ for

388: $-0.09< slope < -0.03$ with rms $= 0.02$. However, to prevent the

389: overestimation of the aperture correction, we restricted the range

390: of the aperture correction to $-0.48< \Delta Mag < 0.0$ (and $-0.1

391: < slope < 0$). The aperture-corrected magnitude is used for

392: deriving the total magnitude in the $V$ band.

393: 

394: \subsection{Size Measurement}

395: \label{size}

396: 

397: We measured the effective radii of the selected clusters using

398: ISHAPE provided by \citet{lar99}. We used the MOFFAT15 model

399: profile for model fit since \citet{lar99} reports that the

400: MOFFAT15 model effectively reproduces the intrinsic half-light

401: radii of the input model star clusters with various model

402: profiles. The model fit was tried to $r = 8$ pixels from the

403: center of the star clusters. The goodness of fit was checked by

404: using the CHISQR value as well as the residual and subtracted

405: images produced by ISHAPE. The resultant FWHM values of the star

406: clusters were transformed into the half-light radii by multiplying

407: by 1.13, according to the guide in the ISHAPE manual.

408: 

409: We derived the empirical point spread function (PSF) necessary for

410: the ISHAPE run by carefully selecting the isolated stars scattered

411: evenly in the neighborhood of NGC 5195. The field around NGC 5194

412: is more densely populated with compact star clusters and it is

413: difficult to avoid contamination by starlike clusters during the

414: PSF candidate star selection. The field of NGC 5195, on the other

415: hand, is not crowded. The possible PSF dependence on the different

416: chips of $HST$ ACS is ignored since NGC 5195 field is already

417: covered by two independent $HST$ ACS pointings where our PSF

418: candidate stars are uniformly distributed.

419: 

420: To test the dependence of the size measurements on the selection

421: of the PSF candidate stars, we ran another ISHAPE operation on

422: about 200 bright star clusters with $V<21$ mag using a new

423: empirical PSF generated from about 20 stars found around the

424: spiral arms of NGC 5194. We found that the difference in the

425: measured sizes of the star clusters is $\Delta R_{\rm eff} < \pm 1$ pc

426: for most clusters: it is found that $\Delta R_{\rm eff}$ does not

427: exceed 3 pc. This result shows that the rate of difference in the

428: measured size is not greater than 20\% for most of the star

429: clusters, and this rate tends to be smaller as the size of the

430: clusters gets larger. For 115 star clusters with $R_{\rm eff} > 2$ pc,

431: which is about 1 pixel wide, the rate of star clusters with

432: $\Delta R_{\rm eff} < 20\%$ is approximately $0.80$.

433: 

434: 

435: \section{Star Cluster Catalog}

436: \label{cat}

437: 

438: We present a catalog of Class 1 star clusters with $V < 23$ mag.

439: Table \ref{gcat} lists a sample catalog of Class 1 star clusters

440: for a reader's guide and the full catalog will be available

441: electronically from the online Journal. Brief descriptions on the

442: data columns of the catalog are given as follows:

443: 

444: Column 1 lists an identification number of a star cluster.

445: 

446: Columns 2 and 3 list the J2000.0 right ascension (RA) and

447: declination (DEC) of a star cluster in degrees.

448: 

449: Columns 4 and 5 list the $V$-band magnitude and corresponding

450: error of a star cluster, measured with an aperture of $r = 6$

451: pixels.

452: 

453: Columns 6 and 7 list the $(B-V)$ color and corresponding error of

454: a star cluster.

455: 

456: Columns 8 and 9 list the $(V-I)$ color and corresponding error of

457: a star cluster.

458: 

459: Column 10 lists the extraction flag of a star cluster given by

460: SExtractor. Flag = 0 means that the star cluster is well isolated.

461: On the other hand, flag = 1 indicates that the cluster has

462: neighbors and flag = 2 indicates that the cluster is originally

463: blended with another source but deblended by SExtractor. If a star

464: cluster has flag = 3, it means that the cluster has a neighbor and

465: is blended with it but deblended by SExtractor.

466: 

467: Column 11 lists the stellarity index of a star cluster. This index

468: has a value between 1 (point sources) and 0 (extended sources).

469: 

470: Column 12 lists the FWHM of a star cluster in arcseconds

471: calculated under the assumption that the cluster has a Gaussian

472: profile.

473: 

474: Column 13 lists the effective radius or half-light radius of a

475: star cluster in pc measured using ISHAPE. See Section \ref{size}

476: for details.

477: 

478: Column 14 lists the ratio of $\chi^2$ defined as

479: $(\chi^2/\chi^2_0)$ calculated by ISHAPE during the effective

480: radius measurement. The $\chi^2$ value is the minimum $\chi^2$

481: calculated after the broadening of the input PSF, while $\chi^2_0$

482: is calculated using the input PSF without any broadening. This

483: ratio indicates how significant the deviation from the input PSF

484: is. The larger this value, the higher the significance of

485: deviation.

486: 

487: Column 15 lists the ellipticity of a star cluster calculated by

488: SExtractor using second-order image moments.

489: 

490: Column 16 lists the aperture-corrected $V$-band magnitude of a

491: star cluster following the method illustrated in Section

492: \ref{apc}.

493: 

494: Column 17 lists any optional comment for a star cluster. If the

495: cluster is classified as a faint fuzzy cluster, then a comment of

496: `FF' is added to this column.

497: 

498: \section{Photometric Properties of Star Clusters}

499: 

500: Figure \ref{cmd} shows the $V$ - $(B-V)$ and $V$ - $(V-I)$

501: color-magnitude diagrams (CMD) of the star clusters in M51. In

502: these diagrams, it can be seen that M51 star clusters, regardless

503: of their classification, are found to be in the range of $-0.2

504: \leq (B-V) \leq 1.5$ and $-0.2 \leq (V-I) \leq 2.0$. The brightest

505: star cluster appears to be as bright as $V \approx 17.5$ ($M_V

506: \approx -12.1$) mag, which is even brighter than $\omega$Cen ($M_V

507: = -10.29$ mag) in our Galaxy. However, most Class 1 and Class 2

508: star clusters in M51 are blue with $(B-V)<0.5$ and $(V-I)<0.8$ and

509: fainter than $V \approx 20$ mag. This suggests that most star

510: clusters in M51 are young. There are some star clusters redder

511: than $(B-V) = 0.5$ and/or $(V-I) = 0.8$. The color spread of Class

512: 1 star clusters is slightly larger than that of Class 2 star

513: clusters. That is, these red star clusters are more abundant in

514: Class 1 than in Class 2.

515: 

516: We derived the $V$-band luminosity function of star clusters,

517: displaying it in Figure \ref{clf}.

518: Figure \ref{clf} shows that the luminosity functions of star clusters

519: in M51, regardless of their classifications, are clearly of

520: power-law form and keep rising up to $M_V \sim -7.0$ mag.

521: The luminosity functions of star clusters become to be slightly flatter for 

522: the fainter magnitudes $M_V \ge -8.0$ mag, 

523: indicating that our cluster detection may be incomplete in this range. 

524: We restricted the magnitude range to $-10.0 <M_V < -8.0$ mag

525: where incompleteness of cluster detection is negligible, 

526: for fitting the data with a power law, $N dL \varpropto L^{\alpha} dL$.

527: It is found that the logarithmic slope $\alpha$ is $-2.59 \pm 0.03$ for all 

528: star clusters in M51 (Fig. \ref{clf}a), while

529: $\alpha = -2.54 \pm 0.04$ for Class 1 clusters (Fig. \ref{clf}b) and

530: $\alpha = -2.70 \pm 0.05$ for Class 2 clusters (Fig. \ref{clf}c),

531: showing only a little difference between the two cluster classes.

532: These slopes of luminosity functions are in good agreement with 

533: the value \citet{gie06b} derived from the WFPC2 data for M51 star clusters

534: for the range of $-11 < M_V < -8$ (see their Figure 7), $\alpha = -2.5 \pm 0.1$.

535: They are also similar to the value for NGC 6946 \citet{lar02} derived from the

536: WFPC2 data  for the range of $-10.75 < M_V < -8.0$, $\alpha = -2.46 \pm 0.12$.

537: 

538: On the other hand, \citet{gie06a} reported that there is a break at $M_V \approx -9.0$, 

539: in the luminosity function of M51 star clusters derived from the same ACS data as used in this

540: study. From fitting with a double power law they derived the slopes, 

541: $-1.93\pm0.03$ for $-6.32>M_V >-8.93$, and 

542: $-2.75\pm0.14$ for $M_V < -8.93$.

543: However, this break is not seen in the luminosity function derived in this study.

544: We calculated $\chi^2$ values for two cases:

545: one with a single power law using the same parameters derived in this study,

546: and the other with a double power law using the same parameters given by \citet{gie06a}.

547: We obtained $\chi^2 = 0.32$ for the double power law fit, which is about twice

548: larger than $\chi^2 = 0.15$ derived from the single power law fit,

549: suggesting that the single power law fit gives or a better fit.

550: 

551: We have investigated any radial dependence of the star clusters'

552: $(V-I)$ color distribution by dividing the observed M51 field into

553: four concentric annuli centered on the NGC 5194 center. Figure

554: \ref{reghist} shows that there is a distinct population of red

555: star clusters at a distance of more than $d = 200 \arcsec$ from

556: the center of NGC 5194, making the $(V-I)$ color distribution for

557: Class 1 clusters clearly bimodal in this region with peaks at

558: $(V-I) \sim 0.5$ and $\sim 1.2$. This indicates that Class 1

559: clusters in this outer region are actually composed of two

560: populations: a blue disk population and a red halo/old disk

561: population, which is also noted in Section \ref{spatdist}. On the

562: other hand, the $(V-I)$ color distributions of the star clusters

563: in the other three annuli display a color peak at $(V-I) \approx

564: 0.5$ and appear to be well represented by a single Gaussian

565: distribution. The number ratio of Class 2 to Class 1 star clusters

566: is about 0.65 on an average for $d < 200 \arcsec$, but it

567: decreases to about 0.28 for $d > 200 \arcsec$.

568: 

569: Figure \ref{ccd} shows the $(V-I)$ vs. $(B-V)$ color-color

570: diagrams (CCD) of Class 1 and Class 2 star clusters. Theoretical

571: evolutionary tracks for the simple stellar population of

572: \citet{bc03} with $Z = 0.02$ and $E(B-V) = 0.1$ for internal

573: reddening are also overlaid on these diagrams. It is evident that

574: most Class 1 and Class 2 star clusters are as old as $10^6 \sim

575: 10^8$ yrs. On the other hand, red star clusters with $(B-V)>0.5$

576: and $(V-I)>0.8$ are considered to be older than $6 \times 10^8$

577: yrs and some of them might be highly reddened younger objects.

578: These old star clusters mostly belong to Class 1 and there are

579: only few old Class 2 clusters. The separation of the population of

580: young and old clusters is seen explicitly for star clusters

581: located in $d>200\arcsec$ from the NGC 5194 center (marked by

582: crosses in Figure \ref{ccd}). That is, young star clusters with

583: $\leq 10^8$ yrs are found on the blue side of $(B-V)=0.5$ and

584: $(V-I)=0.8$ and old star clusters with $\geq 10^9$ yrs are found

585: on the red side. A more detailed quantitative analysis of the star

586: cluster ages using the theoretical population synthesis model will

587: be presented in our forthcoming paper \citep{hwa07b}.

588: 

589: 

590: \section{Spatial Distribution of Star Clusters}

591: \label{spatdist}

592: 

593: Figures \ref{spatc1} and \ref{spatc2} show the spatial

594: distribution of Class 1 and Class 2 star clusters in M51. The most

595: striking impression given by these figures is that star clusters

596: are mostly distributed in and around the two grand design spiral

597: arms of NGC 5194, regardless of their classification. However,

598: Class 2 star clusters are more closely associated with the spiral

599: arms than Class 1 star clusters. As a good example, the width of

600: the southeastern arm of M51, located at about $80 \arcsec$ from

601: the galaxy center, can be compared between the two classes: it is

602: wider for Class 1 than that for Class 2 star clusters by more than

603: double. It is also clear that there are more inter-arm star

604: clusters in Class 1 than in Class 2.

605: 

606: The spatial distribution of Class 1 and Class 2 star clusters

607: shows that the structure of the spiral arms can be traced with

608: star clusters as effectively as the HII regions

609: \citep{ran92a,lee07}, CO or HI gases \citep{rot90,schu07} in

610: galaxies. Therefore, star clusters can be another probe for studying the

611: spiral structure. The spatial distribution of Class 2 star

612: clusters suggests that the visual classification of star clusters

613: based on the morphological shape and the existence of any nearby

614: neighbors is efficient to select young star clusters without

615: utilizing any color information. However, it is necessary to use

616: the color information to select old globular-like star clusters.

617: 

618: Figure \ref{spatc1} shows that Class 1 star clusters with

619: different colors display different spatial distributions. The red

620: clusters with $(B-V)>0.5$ that are expected to be older than $6

621: \times 10^8$ yrs, as shown Figure \ref{ccd} are found to be

622: distributed rather uniformly over NGC 5194 as well as NGC 5195. On

623: the other hand, the blue clusters with $(B-V)<0.5$ are relatively

624: closely associated with the spiral arms of NGC 5194. It also shows

625: that the red clusters in the region of $d>200\arcsec$ are

626: scattered over a wide area around NGC 5195, while the blue

627: clusters in the same region are found only near the spiral arms

628: extended from NGC 5194. However, there is no evident 

629: difference in spatial distribution of the blue and red Class 2 clusters,

630: as shown in Figure \ref{spatc2}. All red Class 2 clusters

631: appear to be associated with the spiral arms, while only some red Class 1 clusters do.

632: This difference in the spatial distribution of the

633: red Class 1 and the red Class 2 clusters suggests that the latter may

634: be highly reddened objects.

635: 

636: The red Class 1 clusters with $(B-V)>0.5$ include very

637: old star clusters with $t\geq 10^9$ yrs as well as intermediate

638: age clusters with $10^8 \leq t < 10^9$ yrs, as suggested in Figure

639: \ref{ccd}. Figure \ref{spatc4} shows the spatial distribution of

640: the very red Class 1 clusters with $(B-V)>0.7$ (panel (d)) as

641: compared to those of the other Class 1 clusters with $(B-V)<0.5$

642: (panel (b)) and $0.5<(B-V)<0.7$ (panel (c)). It appears that the

643: spatial distribution of the star clusters with $0.5<(B-V)<0.7$ is

644: similar to that of the blue clusters with $(B-V)<0.5$, indicating

645: the shape of the spiral arms. On the other hand, it is 

646: difficult to find any spiral-like pattern in the spatial

647: distribution of very red clusters with $(B-V)>0.7$. This suggests

648: that these very red Class 1 clusters with $(B-V)>0.7$ may belong

649: to the old halo of the M51 system. It is also possible that some

650: red Class 1 clusters may be old disk clusters of NGC 5194 and/or

651: NGC 5195.

652: 

653: Figures \ref{spatc3}(a) and (b) show the radial number

654: distributions for Class 1 and Class 2 star clusters. For Class 2

655: star clusters, there are peaks centered at about 3 and

656: 6 kpc, ($74 \arcsec$ and $147 \arcsec$), respectively. The locations of these peaks

657: roughly coincide with the concentric distances of the spiral arms

658: of NGC 5194, which suggests a strong correlation of Class 2

659: clusters with the spiral arms. However, for Class 1 star clusters,

660: the number of clusters rises rather monotonically until $D \approx

661: 6$ kpc. The rapid decrease in the number of clusters in $D > 6$

662: kpc is shared by Class 1 and Class 2 clusters, since this is where

663: the major spiral arms meet their outer boundary. The surface

664: number density of the clusters, shown in panel (b) of Figure

665: \ref{spatc3}, appears to display three distinct parts: a central

666: peak ($D < 2$ kpc), a mid-plateau ($2 < D < 6$ kpc), and an abrupt

667: decline ($D > 6$ kpc). This is in good agreement with the result

668: of \citet{sch06}.

669: 

670: Figures \ref{spatc3}(c) and (d) show the radial

671: distributions of three Class 1 cluster populations with different

672: colors. It is clear that the blue Class 1 clusters with

673: $(B-V)<0.5$ are centrally concentrated within $D = 7$ kpc from the

674: NGC 5194 center, and the surface number density drops sharply at

675: $D \approx 6$ kpc. On the other hand, the red Class 1 clusters

676: with $(B-V)>0.5$ show a rather constant number distribution up to

677: $D \approx 6$ kpc, which leads to a smoothly declining surface

678: number density profile. The very red Class 1 clusters with

679: $(B-V)>0.7$, however, display a flat number distribution across

680: the entire survey field of M51. The surface number density profile

681: of these red clusters shows a small enhancement only at the

682: center, and a flat distribution in the other regions. This may be

683: caused by the preferential disruption of clusters inside the disk.

684: These differences in the spatial and radial distribution of the

685: star clusters, as shown in Figures \ref{spatc1} $\sim$

686: \ref{spatc3}, suggest that the red Class 1 star clusters are

687: composed of old halo population as well as old disk population,

688: survivors of the cluster disruption, and that blue star clusters

689: (both Class 1 and 2) are young disk populations.

690: 

691: 

692: \section{Size Distribution of Star Clusters}

693: 

694: We estimated the sizes of star clusters in M51

695: using ISHAPE and adopted the size measurements with $(\chi^2/\chi^2_0)<0.9$.

696: The ratio $(\chi^2/\chi^2_0)$ indicates how

697: significantly the shape of a source deviates from that of the

698: input PSF. In most cases, however, it suggests the relative

699: extendedness of a source compared to the PSF. The measurements

700: with $(\chi^2/\chi^2_0) \geq 0.9$ mostly returned the radii of

701: $R_{\rm eff}<2$ pc or $1$ pixel, which intrinsically complicates the

702: accurate estimation of size. In particular, in case that the

703: empirically derived PSF is used for the size measurement, as in

704: this study, the implementation of this ratio helps to select only

705: the clusters that return reliable size measurements. However, the

706: adoption of this parameter tends to remove star

707: clusters that have as small radii as $R_{\rm eff}<2$ pc.

708: 

709: Figure \ref{hldist} shows the distribution of the effective radii

710: $R_{\rm eff}$ of star clusters. 

711: The size distributions of Class 1 clusters with $V<23$ (and $V<22$) appear to be composed of three

712: seperate components: a dominant peak at about 3 pc, a weaker one at about 5 pc, 

713: and a long much weaker tail in the larger range.

714: By fitting three Gaussian profiles to the data, we derive

715: the largest component centered at $2.27 \pm 0.03$ pc with $\sigma = 1.90$ pc, %$FWHM = 2.95$ pc,

716: the second component at $4.80 \pm 0.04$ pc with $\sigma = 1.26$ pc, %$FWHM = 1.95$ pc,

717: and the third one at $7.51 \pm 0.30$ pc with $\sigma = 1.38$ pc, %$FWHM = 2.14$ pc, 

718: as shown in Figures \ref{hldist}(a) and (b).

719: The first is very similar to the median size of globular clusters in the Milky Way,

720: which is approximately 3 pc \citep{har96}.

721: The second one represents the intermediate-sized star clusters and

722: they are mostly bluer than $(B-V)=0.5$, as shown in Figure \ref{hldist}(c).

723: The third one corresponds to the large star clusters described in Section \ref{lsc}

724: and it also includes the faint fuzzy clusters \citep{lee05, hwa06}.

725: It is also worth noting that this size distribution of clusters in M51 can be biased against very

726: compact clusters due to the cluster selection criteria and the size measurement constraints adopted. 

727: 

728: Figure \ref{hldist}(c)

729: shows that the blue Class 1 clusters with $(B-V)<0.5$

730: are mostly smaller than $R_{\rm eff}=5$ pc, and their size

731: distribution is consistent with that of the entire cluster sample.

732: However, the red Class 1 clusters with $(B-V)>0.5$ display a large

733: range of radius from $R_{\rm eff} \approx 1$ to $10$ pc or

734: larger and their size distribution may be represented with a

735: very broad single Gaussian centered at $3.66 \pm 0.30$ pc with $\sigma = 3.53$ pc. %$FWHM = 3.50$ pc.

736: One more point to note is that the slope of the size

737: distribution at $R_{\rm eff}>3$ pc region is quite shallower for Class

738: 2 than that for Class 1 clusters. It is the same case with the red

739: Class 1 clusters that display a size distribution slope shallower

740: than that of the blue Class 1 clusters. This suggests that there

741: are a larger number of large Class 2 and red Class 1 clusters.

742: 

743: For globular clusters in the Milky Way and LMC, it is known that

744: there is a correlation between the upper limit of the effective

745: radii ($R_{\rm eff}$) and the galactocentric distance ($D$)

746: \citep{vdb04}. We have investigated any such correlation in M51

747: using Class 1 clusters. 

748: Figure \ref{sizedist} shows the distribution of $R_{\rm eff}$ 

749: vs. $D$ of Class 1 star clusters. 

750: It shows that the effective radii of clusters increase as the

751: galactocentric distance increases, as also suggested by

752: \citet{sch07}. The median $R_{\rm eff}$ of the star clusters located

753: in $D<7$ kpc is about $2.78 \pm 1.64$ pc, but the median $R_{\rm eff}$

754: in $D>8$ kpc is about $4.32 \pm 2.09$ pc. Although the differences

755: in the radii are smaller than the statistical errors, the size

756: distribution of the star clusters in $D < 7$ kpc and in $D > 8$

757: kpc, as shown in the upper panel of Figure \ref{sizedist}, is

758: clearly different. This is in part due to the lack of small

759: clusters in the outer region since the number of star clusters

760: with $R_{\rm eff} < 5$ pc is about 1,500 for $D < 7$ kpc and about 100

761: for $D > 8$ kpc whereas the number of star clusters with $R_{\rm eff}

762: > 7$ pc is about 50 for $D < 7$ kpc and about 30 for $D > 8$ kpc.

763: Moreover, the number of very large star clusters with $10< R_{\rm eff}

764: <20$ pc is about 10 in both the inner and outer regions.

765: Therefore, in the inner region of the spiral arms, there are a large number

766: of typical small star clusters, while large star clusters are

767: evenly distributed over the projected plane of M51.

768: 

769: %The mid and lower panels of 

770: Figure \ref{sizedist}(c) shows the size

771: distribution of the red clusters with $(B-V)>0.5$ and $(V-I)>0.8$ and 

772: and the blue clusters with $(B-V)<0.5$ and $(V-I)<0.8$. These

773: color criteria are consistent with those used to separate Galactic

774: globular clusters. 

775: The size distribution of the red clusters is much broader than that

776: of the blue clusters. %It can be seen clearly that 

777: %The red clusters are relatively larger compared to the blue clusters, which is

778: %consistent with the result shown in Figure \ref{hldist}(c).

779: The median $R_{\rm eff}$ of the red clusters in $D<7$

780: kpc is about $3.45 \pm 2.16$ pc, while that of the blue clusters

781: is about $3.04 \pm 1.59$ pc. Another noteworthy point is that

782: there are many more red clusters with $R_{\rm eff} > 10$ pc than blue

783: clusters. The number of the large red clusters is 16, while that

784: of the large blue clusters is only three.

785: 

786: 

787: \section{Large Star Clusters}

788: \label{lsc}

789: 

790: Apart from typical star clusters with $R_{\rm eff} \approx 3$ pc,

791: large star clusters with $R_{\rm eff} > 7$ pc are being discovered

792: recently: e.g., faint fuzzies in NGC 1023 and 3384

793: \citep{larb00,bro02}, extended star clusters in M31

794: \citep{hux05,mac06}, remote and extended star clusters in NGC 6822

795: \citep{hwa05}, diffuse star clusters in the Virgo cluster

796: \citep{pen06}, and so on. These large star clusters are being

797: considered as a new family of star cluster population. In M51, it

798: has been already reported by \citet{cha04} that there are more

799: large star clusters compared to other nearby late-type galaxies.

800: \citet{lee05} later discovered that there are faint fuzzy clusters

801: around NGC 5195, a companion galaxy of NGC 5194. Based on $HST$

802: ACS data, \citet{hwa06} identified 49 faint fuzzy clusters around

803: NGC 5195 and also reported that these large clusters display a

804: different spatial distribution pattern, compared to the compact

805: clusters with similar colors, implying a different origin for

806: these faint fuzzy clusters.

807: 

808: We define large star clusters (hereafter LSC) as star clusters

809: with $R_{\rm eff}>7$ pc without any reference to the luminosities nor

810: to the colors. The faint fuzzy clusters found in NGC 5195

811: \citep{hwa06} belong to this LSC. The photometric properties of

812: the LSCs are displayed in Figure \ref{lgccmd}. The color-magnitude

813: diagrams in the upper panel of Figure \ref{lgccmd} show that the

814: LSCs are mostly fainter than $V = 21$ mag and only two LSCs are

815: brighter than this. Image inspection revealed that these two

816: bright LSCs are very young clusters located in actively

817: star-forming regions with profuse $H\alpha$ emissions, possibly

818: constituting the so-called star cluster complexes recently

819: reported by \citet{bas05b}. The lower panels of Figure

820: \ref{lgccmd} show the $(B-V)$ and $(V-I)$ color histograms of LSCs

821: in solid lines and Class 1 clusters with $R_{\rm eff}<7$ pc in dashed

822: lines. Please note that the histograms for Class 1 clusters with

823: $R_{\rm eff}<7$ pc are scaled down to 3\% of their real values for a

824: better comparison with those of the LSCs. It is clear that the

825: color distributions for LSCs and Class 1 clusters with $R_{\rm eff}<7$

826: pc exhibit different features: the former displays two distinct

827: peaks, while the latter shows only one blue peak. Another

828: noteworthy point is that there are very few LSCs with $(B-V)<0.1

829: \sim 0.2$ as compared to Class 1 clusters with $R_{\rm eff}<7$ pc.

830: 

831: The $(B-V)$ and $(V-I)$ color distributions of the LSCs shown in

832: the lower panel of Figure \ref{lgccmd} suggest that there are two

833: distinct populations in the LSCs: blue population and red

834: population separated by the colors of $(B-V)=0.5$ and $(V-I)=0.8$.

835: The result of the image inspection suggests that most of the blue

836: LSCs appear to be young clusters associated with other blue

837: clusters. The spatial distribution shown in Figure \ref{lgcspat}(a)

838: also indicates that about half of these blue LSCs

839: are found near the spiral arms of NGC 5194.

840: However, these blue LSCs tend to be found more in the

841: southern region of NGC 5194 disk than in the northern disk.

842: On the other hand, the red LSCs marked by circles in panel (b) of

843: Figure \ref{lgcspat} are found to be more concentrated in the

844: northeastern arm and near the NGC 5194 nucleus than those in the

845: southern disk. Moreover, these red LSCs are widely distributed up

846: to the field of NGC 5195. It is possible that some of these

847: red LSCs in the NGC 5194 disk may be reddened blue LSCs.

848: 

849: The red LSCs include a significant number of faint fuzzy clusters

850: that satisfy the selection criteria defined in \citet{hwa06}.

851: Among the 42 red LSCs, there are 27 faint fuzzy clusters,

852: including 15 previously identified in \citet{hwa06}. These

853: clusters are noted by `FF' comments in the final Class 1 cluster

854: catalog presented in Section \ref{cat}. Figure \ref{ffspat} shows

855: the spatial distribution of these faint fuzzy clusters (panel (b))

856: in comparison with compact red clusters (panel (a)) that are

857: smaller than $R_{\rm eff} = 7$ pc but have the same $(B-V)$ and

858: $(V-I)$ colors as the faint fuzzies. It is clear that the faint

859: fuzzy clusters are more concentrated around NGC 5195 than around NGC

860: 5194. Even the faint fuzzies around NGC 5194 appear to be

861: distributed along the northeast-southwest direction across the

862: disk. Interestingly, this direction is the same as that found for

863: the faint fuzzy clusters associated with NGC 5195 \citep{hwa06}.

864: However, the compact red clusters are distributed rather uniformly

865: over the entire M51 field, which indicates that the compact red

866: clusters and faint fuzzies are different populations.

867: 

868: 

869: \section{Discussion}

870: \label{discuss}

871: 

872: We have detected and selected 2,234 Class 1 and 1,388 Class 2 star

873: clusters in M51 through the visual inspection of about 8,400

874: sources with $V<23$ mag. The final sample of Class 1 star clusters

875: that have a circular shape and no prominent neighbor is compiled

876: into a catalog. This star cluster catalog of M51 has been

877: made using a homogeneous dataset and a consistent selection

878: method over the entire field of M51 including both NGC 5194 and

879: 5195. Unlike the star cluster catalog generated using the

880: automated method, this catalog does not provide a complete list of

881: star clusters in M51. This catalog may suffer from the

882: incompleteness and the subjectivity that are unavoidable in a

883: visual classification process. However, this catalog does provide

884: a least contaminated list of star clusters in M51, which is

885: essential for various studies on star clusters. We have used this

886: cluster catalog and have found some interesting photometric and

887: physical properties of star clusters in M51. This star cluster

888: catalog will be used for a future study on the age distribution of

889: star clusters in M51 \citep{hwa07b}.

890: 

891: We have found that the star clusters in M51 are mostly located

892: near the spiral arms of NGC 5194 regardless of their morphologies.

893: However, Class 2 star clusters appear to be more closely

894: associated with the spiral arms than Class 1 star clusters, some

895: of which are also found between the spiral arms. Another important

896: fact is that a significant number of Class 1 star clusters and

897: nearly all of Class 2 star clusters are bluer than $(B-V)=0.5$ and

898: $(V-I)=0.8$, which suggests that most of the star clusters are

899: young clusters. However, in the outer space of the M51 system,

900: there are more red star clusters with $(B-V)>0.5$ and $(V-I)>0.8$,

901: which is in contrast to the inner disk region of NGC 5194. The

902: $(V-I)$ and $(B-V)$ color-color diagrams show that these red

903: clusters are expected to be old ($t \geq 6 \times 10^8$ yrs). This

904: suggests that even though most star clusters in M51 are blue and

905: young clusters that are formed less than $t \approx 10^8$ yrs ago,

906: there are also some red and old star clusters that should have

907: been formed no less than $t \approx 10^9$ yrs ago, that is, before

908: the epoch of the dynamical interactions between NGC 5194 and 5195.

909: 

910: Old star clusters in M51 are relatively understudied in the

911: previous studies as compared to young and blue clusters. In a

912: recent study on the globular cluster systems of nearby late-type galaxies, the

913: globular cluster specific frequency $S_N$ was calculated by \citet{cha04}. They

914: detected 34 globular clusters in M51 using the $HST$ WFPC2 archive data and

915: calculated the globular cluster specific frequency to be $S_N = 0.5 \pm 0.1$. We

916: assume that the compact red clusters defined in Section \ref{lsc}

917: and shown in Figure \ref{ffspat} are globular clusters, and we

918: calculate the globular cluster specific frequency of M51 using the same method

919: adopted by \citet{cha04}. The number of the compact red clusters

920: with $M_V <-7.2$ mag is 48 and the total number of these clusters

921: would be 96 if we assume that the luminosity function has a

922: symmetric lognormal shape. The corresponding globular cluster specific frequency

923: is $S_N = N_{GC} \times 10^{+0.4(M_V + 15)} \approx 0.20$ if we

924: adopt $M_V = -21.68$ for M51, as used by \citet{cha04}. However,

925: the compact red clusters are a subset of real globular clusters

926: because they are selected based on the same color criteria used

927: for the selection of faint fuzzy clusters, that is, $0.6 < (B-V) <

928: 1.1$ and $1.0 < (V-I) < 1.5$. Therefore, if we relax these

929: criteria and adopt more widely accepted colors for globular

930: cluster selection, that is, $(B-V)>0.5$ and $(V-I)>0.8$, then the

931: number of the selected clusters with $M_V<-7.2$ mag increases to

932: 91, which leads to $S_N \approx 0.39$. This is a rather small

933: value compared to the result of \citet{cha04}. However, we do not

934: apply any type of incompleteness correction to our data. One more

935: thing to consider when we compare these two results is that

936: globular clusters selected by \citet{cha04} include large clusters

937: with $R_{\rm eff}>7$ pc while our definition of globular clusters only

938: allows clusters with $R_{\rm eff}<7$ pc. Including the large clusters

939: into the globular cluster sample would increase the globular cluster specific

940: frequency by about 0.10, yielding $S_N \approx 0.48$, which is in

941: good agreement with the estimation by \citet{cha04}.

942: 

943: The size distribution of star clusters in M51 shows that most of

944: the star clusters selected in this study are as large as $R_{\rm eff}

945: \approx 2.0 \sim 3.3$ pc with a median $R_{\rm eff} = 2.92$ pc. This

946: value is very similar to that of Galactic globular clusters, which

947: is $R_{\rm eff} \sim 3.3$ pc \citep{har96}. However, this size

948: distribution of clusters in M51 can be biased against very compact

949: clusters due to the cluster selection criteria and the size

950: measurement constraints adopted in this study. Recently,

951: \citet{sch07} reported relatively small radii with a median

952: $R_{\rm eff} = 2.1$ pc for their cluster candidates in M51 based on

953: the same Hubble Heritage Data. The cluster size for one extremely

954: large cluster ID212995 noted in their paper (ID54775 in our

955: catalog) also shows a small difference: 21.6 pc in \citet{sch07}

956: and 22.79 pc in this study. The difference in the star cluster

957: size and its distribution between in this study and in

958: \citet{sch07} may be caused by the different cluster selection

959: methods and the different PSFs used in the size measurement.

960: Nonetheless, other characteristics displayed in the size

961: distribution of star clusters still appear to be common in both

962: studies. The logarithmic size distribution of the star clusters

963: between $R_{\rm eff} =3$ and $5$ pc is found to be well represented by

964: the slope of $-1.2$ both in this study and in Figure 14 of

965: \citet{sch07}. As another example, we investigated the size of

966: clusters with $(B-V)<0.1$ and $(B-V)>0.1$ 

967: following the definition by \citet{sch07}. We found that the

968: median $R_{\rm eff}$ for the blue clusters is 2.00 pc and that of the

969: red clusters is 3.24 pc, indicating that the red clusters are

970: systematically larger than the blue clusters. This result is

971: consistent with that of \citet{sch07} which shows that the median

972: $R_{\rm eff}$ is $1.8$ pc for the blue clusters, while it is $2.5$ pc

973: for the red clusters.

974: 

975: The investigation using the star cluster size and the

976: galactocentric distance has returned no obvious correlation

977: between the two parameters, unlike that for Galactic globular

978: clusters. However, the star clusters on the disk of NGC 5194

979: appear to be statistically smaller than the clusters off the disk.

980: This is qualitatively consistent with Figure 19 of \citet{sch07}.

981: Although it is very uncertain due to the large statistical error,

982: it appears that the fraction of large red clusters with

983: $(B-V)>0.5$ and $(V-I)>0.8$ is higher than that of large blue

984: clusters as shown in Figures \ref{hldist} and \ref{sizedist}. For

985: clusters with $R_{\rm eff}>10$ pc, there are 16 large red clusters,

986: while only three such blue clusters are found, which is in

987: striking contrast.

988: 

989: We have found that there are many LSCs

990: in M51. LSCs are mostly fainter than $V = 21$ mag and, unlike

991: typical Class 1 clusters with $R_{\rm eff}<7$ pc, they have two

992: populations with different colors: (1) the blue population with

993: $(B-V)<0.5$ and $(V-I)<0.8$ and (2) the red population with

994: $(B-V)>0.5$ and $(V-I)>0.8$. The blue and red LSCs are found to

995: display slightly different spatial distributions. Many red LSCs

996: are distributed not only over the disk of NGC 5194 but also around

997: the body of NGC 5195, whereas the blue LSCs are mostly located

998: around the NGC 5194 disk. Another interesting point is that even

999: the LSCs around NGC 5194 display different spatial distributions

1000: with different colors: the red LSCs are relatively more

1001: concentrated in the northern disk, while the blue LSCs are

1002: concentrated in the southern disk. It is not yet clear what causes

1003: this difference.

1004: 

1005: Some of these red LSCs are faint fuzzy clusters that satisfy the

1006: color criteria defined in \citet{hwa06}. These faint fuzzy

1007: clusters are preferentially found around NGC 5195, but there are

1008: still some faint fuzzies that appear to be associated with NGC

1009: 5194. Interestingly, the spatial distribution of these faint fuzzy

1010: clusters is very different from that of the compact red clusters,

1011: as shown in Figure \ref{ffspat}. The compact red clusters display

1012: rather uniform distribution over the entire field of M51. On the

1013: other hand, the faint fuzzy clusters are distributed

1014: preferentially along the northeast-southwest direction. In

1015: \citet{hwa06}, the faint fuzzies in NGC 5195 are shown to display

1016: a very similar spatial distribution that is elongated along the

1017: same northeast-southwest direction. It is pointed out by

1018: \citet{hwa06} that the elongated distribution of the faint fuzzies

1019: may be a part of the debris of tidal interactions between NGC 5194

1020: and NGC 5195. If this is the case, the spatial distribution of the

1021: faint fuzzies over the disk of NGC 5194 could be a part of the

1022: same debris. It could provide us with an important clue that might

1023: reveal the passage route of a companion galaxy in the M51

1024: system.

1025: 

1026: 

1027: \section{Summary}

1028: \label{sum}

1029: 

1030: Using the Hubble Heritage Data taken with $HST$ ACS, we 

1031: selected about 3,600 star clusters in M51 based on their

1032: morphological information after visually inspecting about 8,400

1033: sources with $V<23$ mag. We present a final catalog that includes

1034: 2,224 Class 1 star clusters with a circular shape and no prominent

1035: adjacent neighbors. 

1036: %

1037: The selected clusters are mostly found around

1038: the spiral arms of NGC 5194. Most of the star clusters are bluer

1039: than $(B-V)=0.5$ and $(V-I)=0.8$, suggesting that they are young

1040: clusters with $t \leq 10^8$ yrs. However, there are red clusters

1041: around both NGC 5194 and NGC 5195. Some of these red clusters with

1042: $(B-V)>0.7$ and $t \geq 10^9$ yrs are considered to be a part of

1043: the halo population. 

1044: We derived the luminosity function of the clusters, finding that it is

1045: fit well by a single power law 

1046: with $\alpha = -2.59 \pm 0.03$ for the range $-10.0 <M_V < -8.0$ mag.

1047: %

1048: We found that the size distribution of the star clusters can be fit with

1049: three Gaussian components:

1050: a dominant component with a peak at $R_{\rm eff}=2.27 \pm 0.03$ pc with $\sigma = 1.90$ pc, %$FWHM = 2.95$ pc,

1051: a second component with a peak at $4.80 \pm 0.04$ pc with $\sigma = 1.26$ pc, %$FWHM = 1.95$ pc,

1052: and a much weaker broad one with a peak at $7.51 \pm 0.30$ pc with $\sigma = 1.38$ pc.

1053: The first is very similar to the median size of globular clusters in the Milky Way,

1054: %which is approximately 3 pc \citep{har96}.

1055: The second one represents the intermediate-sized star clusters and

1056: they are mostly bluer than $(B-V)=0.5$, %as shown in Figure \ref{hldist}(c).

1057: and the third one corresponds to the large star clusters.

1058: %

1059: Among the large clusters, we found 27 faint fuzzy clusters that are distributed

1060: over the disk of NGC 5194 as well as around the body of NGC 5195.

1061: Interestingly, these faint fuzzy clusters are found to show an

1062: elongated spatial distribution along the northeast-southwest

1063: direction over the NGC 5194 disk. This spatial distribution is

1064: very different from that of the compact red clusters that are

1065: distributed rather uniformly over the entire field of M51. Further

1066: studies on these clusters are needed to understand what induces

1067: this difference as well as the origin of these different star

1068: cluster populations.

1069: 

1070: \acknowledgements

1071: 

1072: The authors are grateful to the anonymous referee for the critical

1073: comments and useful suggestions that helped to improve the original

1074: manuscript.

1075: N.H. and M.G.L. acknowledge the support of the BK21 program of the

1076: Korean Government. This work was supported in part by a grant

1077: (R01-2007-000-20336-0) from the Basic Research Program of the

1078: Korea Science and Engineering Foundation.

1079: 

1080: \begin{thebibliography}{}

1081: %\bibitem[Ashman et al. (1994)]{ash94} Ashman, K. M., Bird, C. M., \& Zepf, S. E. 1994, \aj, 108, 2348

1082: \bibitem[Bastian et al. (2005a)]{bas05a} Bastian, N., Gieles, M., Lamers, H. J. G. L.

1083: M., Scheepmaker, R. A., \& de Grijs, R. 2005a, \aap, 431, 905

1084: \bibitem[Bastian et al. (2005b)]{bas05b} Bastian, N., Gieles, M., Efremov, Y., \& Lamers, H. J. G. L. M.

1085: 2005b, \aap, 443, 79

1086: \bibitem[Bertin \&  Arnouts(1996)]{ber96}Bertin, E., \& Arnounts, S. 1996, \aaps, 117, 393

1087: \bibitem[Mutchler et al. (2005)]{mut05} Mutchler, M. et al. 2005, BAAS, 37, 2

1088: \bibitem[Mackey et al. (2006)]{mac06} Mackey, A. D. et al. 2006, \apj, 653, L105

1089: \bibitem[Bik et al. (2003)]{bik03}Bik, A., Lamers, H. J. G. L. M., Bastian, N., Panagia, N., \& Romaniello, M. 2003, \aap, 397, 473

1090: \bibitem[Brodie \& Larsen (2002)]{bro02}Brodie, J. P., \& Larsen, S. S. 2002, \aj, 124, 1410

1091: \bibitem[Bruzual \& Charlot (2003)]{bc03}Bruzual, \& Charlot, 2003

1092: \bibitem[Chandar et al. (2004)]{cha04}Chandar, R., Whitmore, B., \& Lee, M. G. 2004, \aj, 611, 220

1093: \bibitem[de Vaucouleurs et al. (1991)]{dev91}de Vaucouleurs, G.  et al. 1991, Third Reference Catalog of Bright Galaxies (Berlin:Springer)

1094: \bibitem[Feldmeier et al. (1997)]{fel97}Feldmeier, J. J., Ciardullo, R., \& Jacoby, G. H. 1997, \apj, 479, 231

1095: \bibitem[Gieles et al. (2005)]{gie05} Gieles, M., Bastian, N., Lamers, H. J. G. L.

1096: M., \& Mout, J. N. 2005, \aap, 441, 949

1097: \bibitem[Gieles et al. (2006a)]{gie06a} Gieles, M., Larsen, S. S., Scheepmaker, R. A., Bastian, N., Haas, M. R.,

1098: \& Lamers, H. J. G. L. M. 2006a, \aap, 446, L9

1099: \bibitem[Gieles et al. (2006b)]{gie06b} Gieles, M., Larsen, S. S., Bastian, N., \& Stein, I. T.

1100: 2006b, \aap, 450, 129

1101: \bibitem[Harris (1996)]{har96}Harris, W. 1996, \aj, 112, 1487

1102: \bibitem[Huxor et al. (2005)]{hux05} Huxor, A. P. et al. 2005, \mnras, 360, 1007

1103: \bibitem[Hwang et al. (2005)]{hwa05} Hwang, N., et al. 2005, IAU Colloquium, 198, Near Field Cosmology With Dwarf Elliptical Galaxies, Eds. H. Jerjen \& B. Bingeli (Cambridge: Cambridge Univ. Press), 257

1104: \bibitem[Hwang \& Lee (2006)]{hwa06} Hwang, N., \& Lee, M. G. 2006, \apj, 638, L79

1105: \bibitem[Hwang \& Lee (2007a)]{hwa07a} Hwang, N., \& Lee, M. G. 2007a, IAU Symposium, 241, Stellar Populations as Building Blocks of Galaxies,

1106: Eds. A. Vazdekis, \& R. Peletier (Cambridge: Cambridge University Press), 451

1107: %in press (arXiv:astro-ph/0702011)

1108: \bibitem[Hwang \& Lee (2007b)]{hwa07b} Hwang, N., \& Lee, M. G. 2007b, in preparation.

1109: \bibitem[Larsen(1999)]{lar99}Larsen, S. S. 1999, \aaps, 139, 393

1110: \bibitem[Larsen (2002)]{lar02} Larsen, S. S. 2002, \aj, 124, 1393

1111: \bibitem[Larsen \& Brodie (2000)]{larb00}Larsen, S. S., \& Brodie, J. P. 2000, \aj, 120, 2938

1112: \bibitem[Lee (2003)]{lee03}Lee, M. G. 2003, Journal of Korea Astr. Soc., 36, 189

1113: \bibitem[Lee et al. (2005)]{lee05}Lee, M. G., Chandar, R., \& Whitmore, B. W. 2005, \aj, 130, 2128

1114: \bibitem[Lee (2006)]{lee06}Lee, M. G. 2006, Bulletin of Astr. Soc. India, 34, 99

1115: \bibitem[Lee et al. (2007)]{lee07} Lee, J. H. et al. 2007, in preparation.

1116: \bibitem[Peng et al.(2006)]{pen06} Peng, E. W., et al. 2006, \apj, 639, 838

1117: \bibitem[Rand (1992)]{ran92a}Rand, R. J. 1992, \aj, 103, 815

1118: \bibitem[Rots et al. (1990)]{rot90}Rots, A. H., Bosma, A., van der Hulst, J. M.,  Athanassoula, E., \& Crane, P. C. 1990, \aj, 100, 387

1119: \bibitem[Salo \& Laurikainen (2000a)]{sal00a}Salo, H., \& Laurikainen, E. 2000a, \mnras, 319, 377

1120: \bibitem[Salo \& Laurikainen (2000b)]{sal00b}Salo, H., \& Laurikainen, E. 2000b, \mnras, 319, 393

1121: \bibitem[Scheepmaker et al. (2006)]{sch06}

1122: Scheepmaker, R. A., Gieles, M., Haas, M. R., Bastian, N., Larsen,

1123: S. S., \& Lamers, H. J. G. L. M. 2006, astro-ph/0605022

1124: \bibitem[Scheepmaker et al. (2007)]{sch07}

1125: Scheepmaker, R. A., Haas, M. R., Gieles, M., Bastian, N., Larsen,

1126: S. S., \& Lamers, H. J. G. L. M. 2007, \aap, 469, 925

1127: %(arXiv:0704.3604)

1128: \bibitem[Schlegel et al. (1998)]{sch98} Schlegel, D. J., Finkbeiner, D. P., \& Davis, M. 1998, \apj, 500, 525

1129: \bibitem[Schuster et al. (2007)]{schu07} Schuster, K. F., Kramer,

1130: C., Hitschfeld, M., Garcia-Burillo, S., \& Mookerjea, B. 2007,

1131: \aap, 461, 143

1132: \bibitem[Sirianni et al. (2005)]{sir05}Sirianni, M., et al. 2005, \pasp, 117, 1049

1133: \bibitem[Toomre \& Toomre (1972)]{too72}Toomre, R. P. J., \& Toomre, J. 1972, \apj, 178, 623

1134: \bibitem[van den Bergh \& Mackey (2004)]{vdb04} van den Bergh, S., \& Mackey, A. D. 2004, \mnras, 354, 713

1135: \bibitem[Whitmore et al. (1999)]{whi99}Whitmore, B. C., Zhang, Q., Leitherer, C., Fall, S. M., Schweizer, F., \& Miller, B. 1999, \aj, 118, 1551

1136: \end{thebibliography}

1137: 

1138: \clearpage

1139: 

1140: 

1141: \begin{deluxetable}{rcccccccccccccccc}

1142: \tabletypesize{\scriptsize}

1143: %\tabletypesize{\tiny}

1144: \setlength{\tabcolsep}{0.005in}

1145: \rotate

1146: \tablecaption{A Catalog of Class 1 Bright Star Clusters in M51\label{gcat}\tablenotemark{a}}

1147: \tablewidth{0pt}

1148: 

1149: \tablehead{ \colhead{ID} & \colhead{RA (J2000)} & \colhead{Dec

1150: (J2000)} & \colhead{$V$\tablenotemark{b}} & \colhead{err($V$)} &

1151: \colhead{$(B-V)$\tablenotemark{b}} & \colhead{err$(B-V)$} &

1152: \colhead{$(V-I)$\tablenotemark{b}} & \colhead{err$(V-I)$} &

1153: \colhead{flag} & \colhead{stellarity} & \colhead{FWHM} &

1154: \colhead{R$_{\rm eff}$} & \colhead{$(\chi^2/\chi^2_0)$} &

1155: \colhead{ellipticity} &

1156: \colhead{$V_{apc}$\tablenotemark{b}} & \colhead{remarks} \\

1157:  & \colhead{[deg]} & \colhead{[deg]} & \colhead{[mag]} &  & \colhead{[mag]} &  & \colhead{[mag]} &  &

1158:  &  & \colhead{[arcsec]} &  \colhead{[pc]} & & & \colhead{[mag]} &  }

1159: \startdata

1160:     459 & 202.4433289 &  47.1328163 &  22.146 &   0.015 &   0.873 &   0.027 &   1.223 &   0.018  &    0 & 0.03 &  0.197 &   4.99 &  0.023 &  0.078 &  22.002 &   ... \\

1161:     498 & 202.4610443 &  47.1324654 &  21.258 &   0.010 &   0.328 &   0.015 &   0.794 &   0.013  &    0 & 0.03 &  0.278 &   3.43 &  0.259 &  0.081 &  21.258 &   ... \\

1162:     543 & 202.4734344 &  47.1334114 &  22.967 &   0.024 &   0.833 &   0.041 &   1.193 &   0.028  &    0 & 0.00 &  0.371 &  11.89 &  0.053 &  0.151 &  22.614 &    FF \\

1163:     707 & 202.4062195 &  47.1339684 &  22.023 &   0.014 &   0.234 &   0.021 &   0.524 &   0.020  &    0 & 0.03 &  0.199 &   4.53 &  0.153 &  0.341 &  21.847 &   ... \\

1164:     875 & 202.4083099 &  47.1347275 &  22.173 &   0.016 &   0.132 &   0.022 &   0.363 &   0.022  &    0 & 0.03 &  0.222 &   5.06 &  0.099 &  0.404 &  21.908 &   ... \\

1165:    1084 & 202.4343414 &  47.1359024 &  22.832 &   0.022 &   0.237 &   0.032 &   0.539 &   0.031  &    0 & 0.02 &  0.166 &   3.06 &  0.175 &  0.109 &  22.552 &   ... \\

1166:    1249 & 202.5207825 &  47.1363945 &  21.969 &   0.014 &   0.771 &   0.024 &   1.158 &   0.017  &    0 & 0.07 &  0.151 &   2.55 &  0.057 &  0.049 &  21.936 &   ... \\

1167:    1312 & 202.4645996 &  47.1362038 &  22.199 &   0.016 &   0.325 &   0.024 &   0.930 &   0.020  &    0 & 0.03 &  0.161 &   2.85 &  0.104 &  0.148 &  22.026 &   ... \\

1168:    1366 & 202.4916687 &  47.1367531 &  22.033 &   0.015 &   0.621 &   0.024 &   1.002 &   0.018  &    0 & 0.04 &  0.146 &   2.23 &  0.103 &  0.076 &  21.791 &   ... \\

1169:    1854 & 202.3926697 &  47.1383667 &  22.934 &   0.023 &   0.871 &   0.040 &   1.195 &   0.028  &    0 & 0.02 &  0.170 &   3.50 &  0.079 &  0.008 &  22.676 &   ... \\

1170: %\input{gccat_table3_tex.inp}

1171: \enddata

1172: \tablenotetext{a}{The complete version of this table is in the

1173: electronic edition of the Journal. The printed edition contains

1174: only a sample.} \tablenotetext{b}{All photometric data in this

1175: table are not given in the standard Johnson system but in the

1176: $HST$ equivalents. See Section \ref{phot} for details.}

1177: \end{deluxetable}

1178: 

1179: \clearpage

1180: 

1181: \begin{figure}

1182: \plotone{f1.eps} \caption{ Comparison of photometry of M51

1183: clusters. Boxes represent the boundary for the data used to calculate

1184: the mean offsets that are marked by the sloid lines.

1185: \label{leecomp}}

1186: \end{figure}

1187: 

1188: \begin{figure}

1189:  \begin{center}

1190: \includegraphics[scale=0.65]{f2.eps}

1191:  \caption{Grayscale maps of the representative $V$-band images of a few cluster candidates

1192: overlaid with intensity contours. Panels (a) and (b) show Class 1 cluster candidates, while

1193: panels (c) and (d) show Class 2 cluster candidates. See the text for details.

1194: \label{contour}}

1195:  \end{center}

1196: \end{figure}

1197: 

1198: 

1199: \begin{figure}

1200: \plotone{f3.eps}

1201: \caption{

1202:  Integrated magnitude profiles of isolated bright M51 star clusters.

1203:  A long dashed line marks the size of the aperture we adopted for our aperture

1204:  photometry of star clusters. Note that the magnitude differences (and the slope)

1205:  between at $r=5$ pixels and at $r=7$ pixels (marked by short dashed

1206:  lines) vary from one cluster to another.

1207:  \label{apc1}}

1208: \end{figure}

1209: 

1210: 

1211: \begin{figure}

1212: \plotone{f4.eps} \caption{

1213:  Relationship between the necessary aperture correction $\Delta Mag$ and

1214:  the slope of the integrated magnitude profiles defined at $r=6$ pixels

1215:  for the same isolated star clusters, as used in Fig. \ref{apc1}.

1216:  The solid line shows the linear fit relation between the slope

1217:  and the aperture corrections.

1218: \label{apc2}}

1219: \end{figure}

1220: 

1221: \begin{figure}

1222: \plotone{f5.eps} \caption{

1223:  $V$ - $(B-V)$ and $V$ - $(V-I)$ color-magnitude diagrams of all star

1224:  clusters with $V<23$ mag in M51. The overlaid $(B-V)$ and $(V-I)$ color histograms

1225:  (corresponding y-axes are drawn in the right axis) 

1226:  are for the star clusters with $V<22$ mag. The upper panels (a) and (b) are

1227:  for Class 1 star clusters, while the lower panels (c) and (d) are for Class 2

1228:  star clusters. The typical photometric errors in $(B-V)$ and $(V-I)$ at $V<23$ mag

1229:  do not exceed $0.03$ mag in most cases.

1230: \label{cmd}}

1231: \end{figure}

1232: 

1233: \begin{figure}

1234:  \plotone{f6.eps} \caption{

1235:   Luminosity functions of all (Class 1 and 2) clusters (panel (a)), Class 1 clusters (panel (b)),

1236:   and Class 2 clusters (panel (c)). The faint end cut at $M_V = -6.6$ is caused

1237:   by the criterion of $V<23$ adopted for the star clusters in this study.

1238:   The solid line in each panel indicates the power-law fit of the luminosity function.

1239:  The logarithmic slope $\alpha$ is derived through

1240:   the fit over the range of $-10.0 < M_V <-8.0$.

1241: \label{clf}}

1242: \end{figure}

1243: 

1244: \begin{figure}

1245: \plotone{f7.eps} \caption{

1246:  $(V-I)$ color histograms of M51 star clusters with $V<23$ mag in each annulus

1247:  defined as $d<40\arcsec$ (panel (a)), $40\arcsec<d<120\arcsec$ (panel (b)),

1248:  $120\arcsec<d<200\arcsec$ (panel (c)), and $d>200\arcsec$ (panel (d)), respectively.

1249:  The color distribution of Class 1 star clusters is shown by solid lines,

1250:  while that of Class 2 star clusters is shown by dashed lines.

1251: \label{reghist}}

1252: \end{figure}

1253: 

1254: \begin{figure}

1255: \plotone{f8.eps} \caption{

1256:  $(V-I)$ and $(B-V)$ color-color diagrams of Class 1 star clusters (left panel)

1257:  and Class 2 star clusters (right panel) with $V<23$ mag in M51.

1258:  The crosses represent the star clusters located farther than $200\arcsec$

1259:  from the NGC 5194 center. A theoretical evolutionary track with $Z=0.02$

1260:  by \citet{bc03} is overlaid after applying the foreground reddening of $E(B-V)=0.035$

1261:  and the internal reddening of $E(B-V)=0.1$. Numbers that run from 6.0 to 10.3

1262:  with tick marks on the track represent log(age).

1263:  The arrow in each panel shows the reddening direction.

1264: \label{ccd}}

1265: \end{figure}

1266: 

1267: \begin{figure}

1268: \plotone{f9.eps} \caption{

1269:  Spatial distribution of Class 1 star clusters in M51.

1270:  The large circles in dashed lines indicate the concentric locations

1271:  of $d=40\arcsec$, $120\arcsec$, and $200\arcsec$. The two large crosses

1272:  represent the centers of NGC 5194 and NGC 5195, respectively.

1273:  The blue clusters with $(B-V)<0.5$ are marked by small crosses,

1274:  and the red clusters with $(B-V)>0.5$ are plotted by open

1275:  circles. It is evident that the blue clusters are tightly

1276:  associated with the spiral arms, while the red clusters are distributed

1277:  more uniformly over the entire M51 field.

1278: \label{spatc1}}

1279: \end{figure}

1280: 

1281: \begin{figure}

1282: \plotone{f10.eps} \caption{

1283:  Spatial distribution of Class 2 star clusters in M51.

1284:  The large circles in dashed lines indicate the concentric locations

1285:  of $d=40\arcsec$, $120\arcsec$, and $200\arcsec$. The two large crosses

1286:  represent the centers of NGC 5194 and NGC 5195, respectively.

1287:  The blue clusters with $(B-V)<0.5$ are marked by small crosses,

1288:  and the red clusters with $(B-V)>0.5$ are plotted by open

1289:  circles. It is seen that there are very few red Class 2 clusters

1290:  in M51.

1291: \label{spatc2}}

1292: \end{figure}

1293: 

1294: \begin{figure}

1295:  \plotone{f11.eps}

1296:  \caption{Spatial distribution of Class 1 star clusters with three different $(B-V)$ colors:

1297:   star clusters with $(B-V)<0.5$ (cross) in panel (b), star clusters with $0.5<(B-V)<0.7$

1298:   (open circle) in panel (c), and star clusters with $(B-V)>0.7$ (filled circle) in panel (d).

1299:   Panel (a) shows the spatial distribution of all Class 1 star clusters in dots.

1300:   In each panel, the large circles in dashed lines indicate the concentric locations

1301:  of $d=40\arcsec$, $120\arcsec$, and $200\arcsec$, and the large crosses

1302:  represent the centers of NGC 5194 and NGC 5195, respectively.}

1303:  \label{spatc4}

1304: \end{figure}

1305: 

1306: \begin{figure}

1307:  \plotone{f12.eps}

1308:  \caption{

1309:   Panels (a) and (b): the radial distribution of Class 1 (solid line) and Class 2 (dashed line)

1310:   star clusters in M51 in the number distribution (panel a) and the surface number

1311:   density distribution (panel b). The dotted line represents the number of

1312:   all star clusters (Classes 1 and 2).

1313:   Panels (c) and (d): the radial distribution of the blue Class 1 clusters with $(B-V)<0.5$

1314:   (solid line), the red Class 1 clusters with $(B-V)>0.5$ (slantly shaded), and

1315:   the very red Class 1 clusters with $(B-V)>0.7$ (vertically shaded)

1316:   in the number distribution (panel (c)) and the surface number density distribution

1317:   (panel (d)).

1318:   Please note that the histograms of the red and the very red clusters in panel (c) are

1319:   multiplied by five for a clear presentation.

1320: }

1321: \label{spatc3}

1322: \end{figure}

1323: 

1324: 

1325: \begin{figure}

1326:  \plotone{f13.eps}

1327:  \caption{Panel (a) shows the effective radius $R_{eff}$ distribution

1328:  of Class 1 (solid line) and Class 2 (long dashed line) star clusters

1329:  with $V<23$ mag and $(\chi^2/\chi^2_0)<0.9$ in M51. The dotted line represents

1330:  the distribution of all star clusters (Classes 1 and 2). Panel

1331:  (b) shows the same distribution for the star clusters with $V<22$ mag.

1332:  Gaussian profiles for Class 1 cluster size distribution are plotted in dashed curves

1333:  and the sum of three Gaussians is shown in solid curves.

1334:  Panel (c) shows the size distribution plot for the blue Class 1 clusters

1335:  with $(B-V)<0.5$ (solid line) and the red Class 1 clusters with

1336:  $(B-V)>0.5$ (shaded long dashed line).

1337:  Please note that the histogram of the red Class 1 clusters is

1338:  multiplied by five for a clear presentation.

1339: }

1340: \label{hldist}

1341: \end{figure}

1342: 

1343: 

1344: \begin{figure}

1345:  \plotone{f14.eps}

1346:  \caption{Size variation of the  

1347: Class 1 clusters along the galactocentric distance of NGC 5194. The upper panel shows the

1348: distribution of all Class 1 clusters and the mid panel shows the

1349: same distribution of the red Class 1 clusters with $(V-I)>0.8$ and

1350: $(B-V)>0.5$, while the lower panel displays that of the blue Class

1351: 1 clusters with $(V-I)<0.8$ and $(B-V)<0.5$. In each panel, a

1352: circle represents the mean effective radius of clusters in each

1353: bin, while a triangle represents the median. The histogram in

1354: dashed lines shows the number distribution of Class 1 star clusters along the

1355: galactocentric distance. } \label{sizedist}

1356: \end{figure}

1357: 

1358: \begin{figure}

1359:  \plotone{f15.eps}

1360:  \caption{Upper panels: $(B-V)$ vs. $V$ and $(V-I)$ vs. $V$ color-magnitude diagrams (CMD)

1361: of Class 1 clusters (dots) and LSCs (circles) in M51. Lower

1362: panels: $(B-V)$ and $(V-I)$ color distribution of Class 1 clusters

1363: with $R_{\rm eff} < 7$ pc (dashed lines) and LSCs with $R_{\rm eff} > 7$

1364: pc (solid lines) in M51. Please note that the histograms in dashed

1365: lines display $3\%$ of the real values for a better comparison.

1366: See the text for details.} \label{lgccmd}

1367: \end{figure}

1368: 

1369: \begin{figure}

1370:  \plotone{f16.eps}

1371:  \caption{Spatial distribution of LSCs in M51.

1372: The blue LSCs with $(B-V)<0.5$ are marked by squares in panel (a)

1373: and the red LSCs with $(B-V)>0.5$ are by circles in panel (b),

1374: while Class 1 clusters are plotted in dots in each panel. The

1375: crosses in each panel indicate the centers of NGC 5194 and NGC

1376: 5195, respectively. A large circle with $D=7$ kpc ($d \simeq 172\arcsec$)

1377: is drawn to display the approximate extent of the

1378: disk of NGC 5194. } \label{lgcspat}

1379: \end{figure}

1380: 

1381: \begin{figure}

1382:  \plotone{f17.eps}

1383:  \caption{Spatial distribution of the compact red clusters (panel (a))

1384: and the faint fuzzy clusters (panel (b)) in M51 in comparison with that

1385: of Class 1 clusters plotted in dots. The crosses in each panel

1386: indicate the centers of NGC 5194 and NGC 5195, respectively. A

1387: circle with $D=7$ kpc ($d \simeq 172\arcsec$) is drawn to

1388: display the approximate extent of the disk of NGC 5194.}

1389: \label{ffspat}

1390: \end{figure}

1391: 

1392: 

1393: \end{document}

1394: