1: \documentclass[12pt,preprint]{aastex}
2: %\documentclass[preprint2]{aastex}
3:
4: \begin{document}
5:
6: \title{The intermediate-age globular cluster NGC 1783 in the Large Magellanic Cloud}
7:
8: \author{Alessio Mucciarelli}
9: \affil{Dipartimento di Astronomia, Universit\`a
10: degli Studi di Bologna, Via Ranzani, 1 - 40127
11: Bologna, ITALY}
12: \email{alessio.mucciarelli@studio.unibo.it}
13:
14: \author{Livia Origlia}
15: \affil{INAF - Osservatorio Astronomico di Bologna, Via Ranzani, 1 - 40127
16: Bologna, ITALY}
17: \email{livia.origlia@oabo.inaf.it}
18:
19: \author{ Francesco R. Ferraro}
20: \affil{Dipartimento di Astronomia, Universit\`a
21: degli Studi di Bologna, Via Ranzani, 1 - 40127
22: Bologna, ITALY}
23: \email{francesco.ferraro3@unibo.it}
24:
25: %\author{Flavio Fusi Pecci}
26: %\affil{INAF - Osservatorio Astronomico di Bologna, Via Ranzani, 1 - 40127
27: %Bologna, ITALY}
28: %\email{flavio.fusipecci@oabo.inaf.it}
29:
30:
31: \begin{abstract}
32:
33: We present {\sl Hubble Space Telescope} ACS deep photometry
34: of the intermediate-age
35: globular cluster NGC 1783 in the Large Magellanic Cloud.
36: By using this photometric dataset, we
37: have determined the degree of ellipticity
38: of the cluster ($\epsilon$=0.14$\pm$0.03) and
39: the radial density profile. This profile is well reproduced by a standard
40: King model with an extended core ($r_c$=24.5'') and a low concentration
41: (c=1.16), indicating that the cluster has not experienced the collapse of the core.
42:
43: We also derived the cluster age, by using
44: the {\sl Pisa Evolutionary Library} (PEL) isochrones, with
45: three different amount of {\sl overshooting}
46: (namely, $\Lambda_{os}$=0.0, 0.10 and 0.25).
47: From the comparison of the observed
48: Color-Magnitude Diagram (CMD) and
49: Main Sequence (MS) Luminosity Function (LF) with
50: the theoretical isochrones and LFs,
51: we find that only models with the inclusion of some {\sl overshooting} ($\Lambda_{os}$=0.10-0.25)
52: are able to reproduce the observables.
53: By using
54: the magnitude difference $\delta V_{SGB}^{He-Cl}=0.90$ between
55: the mean level of the He-clump and the
56: flat region of the SGB, we derive an age $\tau$=1.4$\pm$0.2 Gyr.
57:
58:
59: \end{abstract}
60:
61: \keywords{Magellanic Clouds --- globular clusters: individual (NGC~1783) ---
62: techniques: photometry}
63:
64: \section{Introduction}
65: \label{intro}
66:
67: Stellar clusters are key-tracers of stellar populations in different galactic
68: environments. In particular, populous clusters in the Large Magellanic Cloud
69: (LMC) cover a wide range of ages (from a few Myr up to 13 Gyr) which has no
70: counterpart in our Galaxy. Hence, the study of this system allows to extend our
71: empirical knowledge of stellar populations in a mass regime which
72: can be poorly explored in our Galaxy.\\
73: The LMC clusters can be grouped in three
74: main age families, namely: the young population with ages $\le$200 Myr
75: \citep{valle94,testa99}, the intermediate population in the 200 Myr $<$ age $<$
76: 3-4 Gyr range \citep{f95,broc01,gallart03,f04} and the old population, with stellar
77: clusters coeval to the Galactic Halo ones \citep{testa95, brocato96, olsen98,
78: mac04}.\\
79: A few
80: decades ago, the main integrated properties of the LMC cluster system, both in the infrared
81: \citep{persson83} and in the optical \citep{ma79, ma82, swb, vdb81} spectral ranges
82: have been investigated. These studies also provided the only existent homogeneous age-scale,
83: based on the so-called {\sl s-parameter}, as defined by \citet{ef85}. This parameter is an empirical
84: quantity related to the position of the clusters in the (U-B)-(B-V) color-color diagram.
85: Clearly, this method presents many uncertainties, namely the
86: foreground/background contamination and the possible statistical fluctuations due
87: to bright stars.
88:
89: The advent of 8-meter class ground-based
90: telescopes and the superior performances of the Hubble Space Telescope (HST) provide
91: sufficient resolution
92: to properly study these clusters even in their innermost crowded regions.
93: Accurate ages can be determined from the Main
94: Sequence (MS) Turn-Off (TO) measurements \citep[see e.g. the recent works
95: by][]{mac06,kerber,m07}, once updated theoretical evolutionary models are
96: adopted and precise estimates of the global metallicity \citep{salaris} be
97: available. Indeed, the stellar clock is extremely sensitive to the chemical
98: composition and detailed abundances of iron and $\alpha$-elements from
99: high-resolution spectroscopy are mandatory for this purpose.
100:
101: A few years ago, we started a long term project aimed at determining homogeneous ages and
102: metallicities for a representative sample of template LMC clusters, by combining
103: high-resolution photometry and spectroscopy. The first cluster analyzed so far is
104: NGC 1978: an accurate metallicity of [Fe/H]=-0.37$\pm$0.07 dex \citep{f06} and an
105: age of $\tau$=1.9$\pm$0.1 Gyr \citep[][hereafter Paper I]{m07} have been obtained. In
106: this paper we present the results for NGC~1783, another populous intermediate-age cluster.
107: Sect.~\ref{obs} describes the cluster Color-Magnitude Diagram (CMD) and
108: its main evolutionary features. Sect.~\ref{parsec} describes its structural parameters,
109: while
110: Sect.~\ref{agesec} discusses its age determination.
111: In Sect.~\ref{disc} we draw our conclusions.
112:
113:
114: \section{Observations and data analysis}
115: \label{obs}
116:
117: The results presented in this paper are based on a set of images obtained
118: with the Advanced Camera for Survey (ACS) Wide Field Channel (WFC) that
119: provides a field of view of $\approx$200'' $\times$200'' with a
120: plate scale of 0.05 arcsec/pixel. All the images have been retrieved
121: from the ESO/ST-ECF Science Archive (Proposal ID 9891, Cycle 12),
122: through the F555W and F814W filters, with
123: exposure times of 250 and 170 sec, respectively.
124: The first chip of the ACS-WFC
125: is centered on the cluster center.
126: Fig.~\ref{fits} shows the
127: the F814W image of the cluster in both the ACS chips.\\
128: The photometric reduction was carried out with the
129: {\it DAOPHOT-II} package \citep{stet} by using the
130: Point Spread Function (PSF) fitting method.
131: The final photometric catalog includes almost 40,000 stars,
132: and it has been calibrated in the VEGAMAG photometric system using the
133: prescriptions of \citet{bedin} and astrometrized on the
134: Two-Micron All-Sky Survey (2MASS) photometric system,
135: by cross-correlating the ACS@HST catalog with the infrared catalog
136: presented by \citet{m06}.
137:
138: \subsection{The CMD overall characteristics}
139: \label{cmdsec}
140:
141: Fig.~\ref{cmd} shows the observed CMD
142: using only the ACS chip sampling the cluster core.
143: The useful magnitude range is $\rm 17.6\le F555W \le 26$.
144: Indeed, we note that the brightest stars
145: at F555W$<$17.6, could
146: be in the non-linear regime of the CCD or saturated in their central pixels,
147: making the
148: corresponding magnitudes and colors somewhat uncertain.\\
149: The main features of the observed CMD can be summarized as follows:\\
150: (1) The MS extends over more than 6 magnitudes in the
151: F555W band and the TO point is located at F555W$\approx$21.2
152: (the identification of the TO magnitude was done by means of a
153: parabolic fit of this region).
154: The TO region shows a mild spread in color;\\
155: (2) the slope change of the MS is at F555W$\approx$22.2 and
156: flags the transition between radiative and convective core stellar structures;\\
157: (3) the Sub Giant Branch (SGB) is a poorly populated sequence,
158: with a typical F555W$\approx$20.5 magnitude.
159: We note that the blue edge of this sequence is not well-defined;\\
160: (4) the Red Giant Branch (RGB) is well populated and it extends
161: over $\approx$5 magnitudes;\\
162: (5) the Helium-Clump is located at F555W$\approx$19.25 and
163: (F555W-F814W)$\approx$1.15; \\
164: (6) the Asymptotic Giant Branch (AGB) Clump
165: (corresponding to the base of the AGB sequence)
166: is visible at F555W$\approx$18.4.
167:
168: Fig.~\ref{radcmd} shows the radial
169: CMDs by using the entire sample of stars detected in the ACS FoV.
170: The bulk of the cluster population lies in the central 2 arcmin
171: (by radius); at r$>$130'' the SGB,
172: RGB and He-Clump are barely detectable, while the brightest portion of the
173: cluster MS is still visible.
174:
175: The mild color broadening of the TO region deserves a brief discussion.
176: Recently, \citet{ber} found a color dispersion in the brightest
177: portion of the MS of NGC 2173, while \citet{mac07} found
178: a bifurcation of the bright MS region of NGC 1846, and interpreted it
179: as a double TO.
180: These two observational evidences suggest the possible existence
181: of an age-dispersion in these stellar clusters.
182: In order to check whether the broadening of the TO region in NGC 1783 can be
183: ascribed to a possible age-dispersion as well,
184: we calculated the color distribution of the MS stars in the 20.5$<$F555W$<$21.1 magnitude range.
185: The color distribution turns out to be
186: roughly Gaussian with $\sigma_{F555W-F814W}\approx$0.05, which is fully
187: consistent with the observational errors ($\sigma_{F555W}\sim\sigma_{F814W}\approx$0.03,
188: implying a color uncertainty $\sigma_{F555W-F814W}\approx$0.04). Similar results
189: are obtained by computing the color distribution in the radial CMDs of Fig.~\ref{radcmd}.
190: Thus, we can conclude that the spread in color of the TO region in NGC 1783 can
191: be explained in terms of photometric errors and there is not any
192: evidence of an age-dispersion.
193:
194: \subsection{Completeness}
195: \label{compsec}
196:
197: In order to quantify the degree of completeness of the final photometric catalog,
198: we used the well-know artificial star technique \citep{mateo}, and we simulated a
199: population of stars in the same magnitude range covered by the observed CMD
200: (excluding stars brighter than F555W=17.6, corresponding to the saturation level) and
201: with a (F555W-F814W)$\sim$0.8 mean color. The artificial stars have been added to the
202: original images and the entire data reduction procedure has been repeated using the
203: {\sl enriched} images. The number of artificial stars simulated in each run ($\sim$
204: 2,000) is always a small percentage ($\sim$5\%) of the detected stars, hence they do
205: not alter the original crowding conditions. A total of $\sim$250 runs were performed
206: and more than 500,000 stars have been simulated.
207: We have excluded from our analysis the very inner
208: region of the cluster (r$<$20''), where the crowding conditions are prohibitive.
209: Fig.~\ref{comp} shows the completeness factor $\phi=\frac{N_{rec}}{N_{sim}}$, defined as the
210: fraction of recovered stars over the total simulated ones, as a
211: function of the F555W magnitude in two different radial regions, namely between 20''
212: and 50'' and at r$>$50'' from the cluster center, respectively.
213: In the inner region the sample is $>$90\% complete down to F555W$\approx$22.5,
214: while in the outer region is $>$90\% complete down to F555W$\approx$24.
215:
216: \section{Ellipticity and structural parameters}
217: \label{parsec}
218:
219: The knowledge of the position of each star over the entire
220: extension of the cluster (and in particular in the innermost region)
221: allows to compute the center of gravity ($C_{grav}$)
222: with high precision. In doing this, we applied the procedure
223: described in \citet{mon95}, averaging the $\alpha$ and $\delta$
224: coordinates of the detected stars with F555W$<$22,
225: in order to minimize the effects of
226: incompleteness.
227: The $C_{grav}$ of
228: the cluster turns out to be located at $\alpha$=$4^h$ $59^m$ $09^s$.78
229: and $\delta$=-$65^{\circ}$ 59' 17''.82.
230: This finding is in good agreement with our previous
231: determination based on
232: near-IR photometry \citep{m06}.
233:
234: We also used the ACS photometry of NGC 1783 to derive new estimates for the cluster
235: ellipticity and structural parameters. The isodensity curves are computed
236: with an adaptive kernel technique, accordingly to the prescription of
237: \citet{fuk}. We used all the stars in the first chip with F555W$<$22 in order to
238: minimize incompleteness effects and we fit the isodensity curves with ellipses.
239: Fig.\ref{isoden} shows the cluster map with the isodensity contours (upper panel), the
240: corresponding best fit ellipses (central panel) and their ellipticity as a function
241: of the semi-major axis in arcsec (lower panel). The ellipticity $\epsilon$
242: (defined as $\epsilon$=1-(b/a), where a and b are the major and minor axis of the
243: ellipse, respectively) turns out to be 0.14$\pm$0.03. This value results slightly
244: lower than the previous determinations of \citet{geisler} that found an average
245: ellipticity of $\epsilon$=0.19.
246:
247: By following the procedure already described in previous
248: papers \citep[see][]{f04b},
249: we also compute the projected density
250: profile of the cluster.
251: The area sampled by the first ACS chip has been divided
252: in 18 concentric annuli, each one centered on $C_{grav}$ and split in four sub-sectors.
253: The number of stars lying in each sub-sector was counted and the mean star density
254: was obtained. The standard deviation was estimated from the variance among the
255: sub-sectors. The radial density profile is plotted in ~Fig.\ref{radprof}. \\ We used the
256: \citet{sig} code in order to compute the family of isotropic single-mass King
257: models. These models are defined by three main parameters, the central potential
258: $W_0$, core radius $r_c$ and the concentration
259: c=$\log{(r_t)/(r_c)}$, where $r_t$ is the tidal radius.
260: Fig.~\ref{radprof} also shows the single-mass
261: King model that best fit the derived density profile. The best-fit model has been
262: selected
263: by using a $\chi^2$ minimization (shown in the lower panel of Fig.~\ref{radprof}).
264:
265: We find
266: $W_0$=5.5, $r_c$=24.5'' and c=1.16 , corresponding to a tidal radius $r_t$= 5.9'
267: \footnote{We underline that the structure of the profile and the corresponding
268: derived parameters does not change if different magnitude limits are adopted.}.
269: Our estimate of $r_c$ is consistent with the one by \citet{elson92} who found
270: $r_c$=20''.
271: The resulting $r_t$ lies out of the field of view of ACS.
272: In order to properly fit the most external points
273: of the radial profile,
274: the best-fit King model has been combined with a constant background level
275: (corresponding to a density of 350 stars $arcmin^2$), and shown as a
276: horizontal dashed line
277: in Fig.~\ref{radprof}.
278:
279: \section{The age of NGC 1783}
280: \label{agesec}
281:
282: Young stellar populations (with ages $\le$300 Myr) are characterized by large
283: convective cores.
284: Theoretical studies \citep[see e.g. the numerical
285: simulations computed by][]{freytag} suggest that
286: the penetration of convective
287: elements into a stable region ({\sl via the Schwarzschild criterion})
288: can produce non-negligible evolutionary effects.
289: These prediction seem to be confirmed by several works
290: \citep{bm83,barmina,chiosi07} which require
291: some amount of {\sl
292: overshooting} in the MS star convective core , in order
293: to reproduce the observed morphologies
294: and stellar counts of young clusters, although this issue
295: is still matter of
296: debate \citep{testa99,broc03}.
297: At variance, in older ($\ge$5-6 Gyr) stellar populations the
298: growth of large radiative cores tends to erase the possible evolutionary effects
299: of {\sl overshooting}.
300:
301: Intermediate-age stellar populations like those in NGC 1978 and
302: NGC 1783 LMC stellar clusters represent the transition stage between
303: these two regimes, and
304: thus represent ideal test-bench to study the {\sl overshooting} effects.
305:
306: \subsection{Basic assumptions}
307: \label{basic}
308:
309: In Paper I we
310: performed a detailed
311: comparison of the observed morphology and star counts
312: of NGC 1978 with different set of theoretical
313: models and {\sl overshooting} efficiencies.
314: The best agreement between observations and theoretical predictions was
315: reached with the Pisa Evolutionary Library (PEL).
316: \footnote{The PEL isochrones are available at the URL
317: http://astro.df.unipi.it/SAA/PEL/Z0.html.}
318:
319: Hence, we have used the PEL isochrones to also determine the age of NGC 1783.
320: We select isochrones with Z=0.008 (corresponding to
321: [M/H]=-0.40 dex, as estimated by Mucciarelli et
322: al.(2007, in preparation) from high-resolution spectroscopy), and with three
323: different amount of {\sl overshooting} efficiency, namely
324: $\Lambda_{os}$
325: \footnote{The overshooting efficiency is parametrized
326: using the mixing length theory
327: \citep{bome} with $\Lambda_{os}$=1/$H_p$
328: (where $H_{p}$ is pressure scale height) that quantifies the overshoot distance {\sl above}
329: the Schwarzschild border in units of the pressure scale height.}
330: =0.0 for the canonical isochrones, and $\Lambda_{os}$=0.10
331: and 0.25, representative of mild and strong {\sl overshooting} regimes, respectively.
332:
333: These theoretical isochrones have been transformed into the observational plane,
334: by means of suitable conversions computed with the code described
335: by \citet{origlia}, and convolving the model atmospheres by \citet{bcp}
336: with the ACS filter responses. Guess values
337: of $(m-M)_0$=18.50 \citep{alves} for the distance modulus and
338: E(B-V)=0.10 \citep{persson83} for reddening have been adopted. However,
339: in order to obtain
340: the best fit of the observed sequences we allowed these parameters
341: to vary by $\le|10|$\% and $\le|40|$\% factors, respectively.
342:
343: Fig.~\ref{isoc} shows the best-fit solutions
344: for the different values
345: of $\Lambda_{os}$, as obtained by
346: matching the following features:\\
347: (1) the magnitude of the He-Clump;\\
348: (2) the magnitude difference between the He-Clump and the flat
349: region of the SGB;\\
350: (3) the difference in color between the TO
351: and the base of the RGB.
352:
353: As can be seen, the canonical model with $\Lambda_{os}$=0.0 fit the
354: observational features (1) and (2) reasonably well
355: with $(m-M)_0$=18.57, E(B-V)=0.13 and
356: $\tau$=0.9 Gyr,
357: but fails to reproduce feature (3).
358:
359: Fig.~\ref{isoc2} (panel (a)) shows a portion of the CMD, as zoomed onto the TO region,
360: with the best-fit ($\tau$=0.9 Gyr) and 0.3 Gyr older ($\tau$=1.2 Gyr) isochrones.
361: The older isochrones better fits feature (3) but predicts a too bright
362: (by $\approx$0.3 magnitudes) He-clump.
363: Moreover, it requires a $(m-M)_0$=18.16 distance modulus, which is
364: definitely too short for the LMC \citep{alves}.\\
365: Fig.~\ref{isoc2} (panels (b) and (c)) shows
366: a similar comparison for the overshooting models.
367: For the $\Lambda_{os}$=0.10 model (panel (b)), the best-fit ($\tau$=1.2 Gyr) and 0.2 Gyr older
368: ($\tau$=1.4 Gyr) isochrones are plotted.
369: As for the canonical model,
370: the older isochrones somewhat better fits feature (3) but predicts a too bright
371: (by $\approx$0.25 magnitudes) He-clump and a too short
372: $(m-M)_0$=18.25 distance modulus.\\
373: For the $\Lambda_{os}$=0.25 model (panel (c)), the best-fit ($\tau$=1.6 Gyr) and 0.2 Gyr younger
374: ($\tau$=1.4 Gyr) isochrones are shown.
375: The younger isochrone sligthly better fits the SGB region but predicts a too blue MS.
376: Also it predicts a sligthly too faint (by $\approx$0.2 magnitudes) He-clump
377: and too long $(m-M)_0$=18.66 distance modulus.\\
378: In summary, we can conclude that canonical models,
379: regardless the adopted isochrone age,
380: do not provide an acceptable fit
381: to the observed CMD, while models with
382: $\Lambda_{os}$=0.10 and 0.25 {\sl overshooting},
383: E(B-V)=0.13, $(m-M)_0$=18.45
384: and ages between $\tau$=1.2 and $\tau=$1.6 Gyr, respectively,
385: reasonably well reproduce all the three diagnostics features.
386:
387: \subsection{Star counts and overshooting efficiency}
388: \label{count}
389:
390: A quantitative check to
391: discriminate between the different {\sl overshooting scenarios} is to
392: perform a comparison between the observed and theoretical LFs
393: of the MS stars normalized to the number of the He-clump stars, defined as
394: $$\Phi_{norm}=\lg{\frac{\sum_{i}{N_{MS}}}{N_{He-Cl}}}.$$
395: Such a normalized LF is a powerful indicator of the relative timescales of the H and He burning
396: phases.
397: The observed $\Phi_{norm}$ is obtained by counting the number of MS stars
398: ($N_{MS}$) in each 0.5 magnitude bin, after the
399: correction for incompleteness and field contamination, and normalized
400: to the total number of He-Clump stars.
401: The innermost region of the cluster ($r<20''$, see Fig.~\ref{radcmd}) has been
402: excluded from this analysis
403: because of its prohibitive crowding.
404: Formal errors for the observed $\Phi_{norm}$ in each magnitude bin
405: are computed under the assumption that star counts
406: follow the Poisson statistics,
407: by using the following formula:
408: $$\sigma_{\Phi_{norm}}=\frac{\sqrt{{\Phi_{norm}}^{2}\cdot\sigma_{N_{He-Cl}}^{2}+\sigma_{N_{MS}}^{2}}}{N_{He-Cl}}.$$
409:
410: Since the ACS field of view is not large enough to properly sample
411: the field population around NGC 1783,
412: we used the most external region ($r>150''$) of the
413: decontamination field for NGC1978 (Paper I).
414: Indeed, these two clusters are close enough for the purpose of decontamination and
415: their field RGB sequences are well-overlapped.
416:
417: Fig.\ref{dec} shows the histogram of the number of MS stars per $arcmin^{2}$
418: at r$>$150'' from the center of
419: NGC 1978.
420: The number of MS and He-Clump stars in this field have been
421: subtracted from the NGC 1783 cluster stellar counts,
422: after the normalization for the sampled area.
423:
424: Hence, the total number of stars in
425: each magnitude bin is given by:
426: $$N_{corr}=\frac{N_{obs}}{\phi}-N_{field}.$$
427:
428: In order to compute the theoretical $\Phi_{norm}$ predicted by the PEL models, we have
429: adopted the well-know technique of synthetic diagrams. By using the
430: best-fit models described above, we randomly distributed the stars
431: along the isochrone accordingly to a Salpeter initial mass function.
432: An artificial
433: dispersion has been added in order to simulate the photometric errors.
434: For each model, 200
435: synthetic diagrams are computed by using Montecarlo simulations, and
436: the corresponding $\Phi_{norm}$ are extracted
437: and averaged together. \\
438:
439: Fig.~\ref{lumfun} (panel {\sl (a)}) shows the observed LF (black points)
440: compared with the theoretical expectations,
441: computed by using the three different {\sl overshooting} models.
442: Clearly, the $\Lambda_{os}$=0.0 model
443: predicts a $\Phi_{norm}$ value $\sim$10-15\% lower than the observed one;
444: the $\Lambda_{os}$=0.10 and 0.25 models marginally ($<$5\%)
445: underestimate the observed value of $\Phi_{norm}$.
446: This small offset can be easily accounted for by adding a binary population
447: in the synthetic LF. To do this, we assumed that a
448: given fraction {\sl $f_b$} of the simulated stars be
449: the primary star of a binary system.
450: The mass of the
451: primary is randomly extracted, while
452: the mass of the secondary star
453: is assigned by adopting the mass ratio {\sl q},
454: between the secondary and primary star.
455: The magnitude of the binary system is given by
456: $M_{F555W}^{Binary}=-2.5\cdot\log{(10^{-2.5\cdot(M_{F555W}^{prim}+M_{F555W}^{sec})}})$,
457: where $M_{F555W}^{Binary}$, $M_{F555W}^{prim}$ $M_{F555W}^{sec}$ are the magnitudes
458: of the binary, the primary and the secondary star, respectively.
459: The latter has been obtained from the isochrone mass/luminosity relation.
460: Panel {\sl (b)} in Fig.~\ref{lumfun} shows the comparison between
461: the observed and theoretical $\Phi_{norm}$
462: with a binary population.
463: The inclusion of $\approx$10\% binaries with a flat
464: distribution of mass ratios ({\sl q}=0.80) provide a good match
465: between theoretical and observed $\Phi_{norm}$ for the models with
466: overshooting. A residual discrepancy of $\approx$10\% is
467: still present between the observed and the theoretical $\Phi_{norm}$ as predicted
468: by the
469: $\Lambda_{os}$=0.00 model.
470: The adopted binary fraction is
471: somewhat smaller than previous estimates ($\le 30$\%) in other LMC and SMC
472: clusters \citep{testa95,barmina,chiosi07}.
473:
474: \section{Discussion and Conclusions}
475: \label{disc}
476:
477: The overall CMD characteristics of NGC 1783 are quite similar
478: to those of NGC 1978 (Paper I), although there is evidence
479: of an age difference. Indeed,
480: we have shown that
481: the best fit solutions to the observed CMD features are obtained
482: by selecting
483: $\Lambda_{OS}$=0.1-0.25 and $\tau$=1.2-1.6 Gyr
484: for NGC 1783 (see Sect.~\ref{agesec}) and
485: $\Lambda_{OS}$=0.1 and $\tau$=1.9 Gyr for NGC 1978 (Paper I).
486:
487: Further insight on the relative age of the two clusters can be obtained from the direct
488: cluster-to-cluster comparison of the overall CMD properties. To this aim
489: we can define the $\delta V_{SGB}^{He-Cl}$ parameter as the magnitude difference
490: between the luminosity distribution peak of the He-Clump and the flat region of
491: the SGB. This {\it differential} parameter
492: can provide an independent estimate of the age,
493: and it is formally the analogous of the so-called {\sl vertical method},
494: based on the
495: magnitude difference between the TO and the Horizontal Branch magnitude level,
496: and used to infer the age for the old globulars \citep[see e.g.][]{buo89}.
497:
498: Fig.~\ref{comcmd} shows the two observed CMDs with marked the
499: $\delta V_{SGB}^{He-Cl}$ parameter:
500: we find $\delta V_{SGB}^{He-Cl}=0.90$ and $1.56$ for
501: NGC1783 and NGC1978, respectively. This difference is an independent, clearcut
502: indication that NGC 1783 is younger than NGC 1978.
503:
504: Fig.~\ref{theo} shows the
505: theoretical relations between the $\delta V_{SGB}^{He-Cl}$ observable and the age,
506: as derived from the PEL models with different amounts of {\sl overshooting}.
507: The grey area marked the region of the ($\tau$,
508: $\delta V_{SGB}^{He-Cl}$) plane for a mild/strong overshooting efficiency
509: appropriate for NGC 1783.
510: Hence, by entering the measured $\delta V_{SGB}^{He-Cl}$ in the above relations,
511: an independent estimate of the age based on this differential
512: parameter can be obtained.
513:
514: By using the measured value of $\delta V_{SGB}^{He-Cl}=0.90$,
515: we find $\tau$=1.4$\pm$0.2$\pm$0.1 Gyr for NGC 1783,
516: where the first errorbar
517: refers to the uncertainty in {\sl
518: overshooting} efficiency and the second to the uncertainties in
519: the adopted reddening and distance modulus.\\
520:
521: This age is still consistent with the
522: one inferred by \citet{geisler97} ($\tau$=1.3 Gyr), while it is significantly
523: older than the age derived from the s-parameter ($\tau\sim$0.9 Gyr) and
524: by \citet{mould} ($\tau$=0.7-1.1 Gyr).
525: \citet{m06} note that
526: the $N_{Bright-RGB}$/$N_{He-Cl}$ population ratio computed for NGC 1783
527: is too high for the clusters undergoing the RGB Phase-Transition, as
528: suggested by the s-parameter age. Our new determination of an older age for
529: NGC 1783, better reconcile the $N_{Bright-RGB}$/$N_{He-Cl}$ population ratio
530: with the observed well-populated RGB.
531:
532: Finally, we note that
533: the structural parameters ($r_c$, $r_t$) and the age
534: of the cluster
535: inferred from this study, allow us to
536: constrain the dynamical state of this cluster. The resulting core radius of
537: $r_c$=24.5'' (corresponding to $\sim$5.9 pc adopting the distance modulus of
538: $(m-M)_0$=18.45, obtained from the best-fit with the {\sl overshooting} models (see
539: Sect.~\ref{agesec})) is consistent with the age-core radius relationship discussed by
540: \citet{mac03} and based on the surface brightness radial profiles of 53 LMC rich
541: clusters. The youngest (ages $<\sim$200 Myr) clusters of their sample exhibit
542: core radii $<$3 pc, while the older
543: (both intermediate and old-age)
544: stellar clusters
545: show a more scattered distribution, with $r_c$ between $\sim$1 and $\sim$8
546: pc, a major peak at $r_c\sim$2.5 pc and the presence of several objects with
547: $r_c>\sim$5 pc. \\ The inferred concentration parameter, c=1.16, is consistent with
548: a not core-collapse cluster \citep{meylan97}, as expected given the relatively
549: young age of NGC 1783.
550:
551:
552: \acknowledgements
553:
554: This research
555: was supported by the Agenzia Spaziale Italiana (ASI) and the
556: Ministero dell'Istruzione, del\-l'Uni\-versit\`a e della Ricerca.
557:
558:
559: \begin{thebibliography}{}
560: \bibitem[Alves(2004)]{alves}
561: Alves, D. R., 2004, New Astronomy Review, 48, 659
562: \bibitem[Barmina, Girardi \& Chiosi(2002)]{barmina}
563: Barmina, R., Girardi, L., \& Chiosi, C., 2002, \aap, 385, 847
564: \bibitem[Becker \& Mathews(1983)]{bm83}
565: Becker, S., \& Mathews, J., 1983, \aj, 270, 155
566: \bibitem[Bedin et al.(2005)]{bedin}
567: Bedin, L. R., Cassisi, S., Castelli, F., Piotto, G., Anderson, J., Salaris, M., Momany, Y. \& Pietrinferni, A.,
568: 2005, \mnras, 357, 1048
569: \bibitem[Bertelli et al.(2003)]{ber}
570: Bertelli, G., Nasi, E., Girardi, L., Chiosi, C., Zoccali, M., \&
571: Gllart, C., 2003, \aj, 125, 770
572: \bibitem[Bessel, Castelli \& Plez(1998)]{bcp}
573: Bessel, M. S., Castelli, F., \& Plez, B., 1998, \aap, 333, 231
574: \bibitem[Bohm-Vitense (1958)]{bome}
575: Bohm-Vitense, E., 1958, ZA, 46, 108
576: \bibitem[Brocato et al.(1996)]{brocato96}
577: Brocato, E., Castellani, V., Ferraro, F. R.,Piersimoni, A. M.,
578: \& Testa, V.,1996, \mnras, 282, 614
579: \bibitem[Brocato et al.(2001)]{broc01}
580: Brocato, E., Di Carlo, E., \& Menna, G., 2001, \aap, 374, 523
581: \bibitem[Brocato et al.(2003)]{broc03}
582: Brocato, E., Castellani, V., Di Carlo, E., Raimondo, G., \& Walker, A. R.,
583: 2003, \aj, 125, 3111
584: \bibitem[Buonanno, Corsi \& Fusi Pecci(1989)]{buo89}
585: Buonanno, R., Corsi, C. E., \& Fusi Pecci, F. 1989, \aap, 216, 80
586: \bibitem[Castellani et al.(2003)]{cast03}
587: Castellani, V.,Degl'Innocenti, S., Marconi, M., Prada Moroni, P.G., Sestito, P.,
588: 2003, \aap, 404, 645
589: \bibitem[Chiosi \& Vallenari(2007)]{chiosi07}
590: Chiosi, E., \& Vallenari, A., 2007, \aap, astro-ph/070228
591: \bibitem[Elson \& Fall(1985)]{ef85}
592: Elson, R. A., \& Fall, S. M. 1985, \apj, 299, 211
593: \bibitem[Elson (1992)]{elson92}
594: Elson, R. A., 1992, \mnras, 256, 515
595: \bibitem[Ferraro et al.(1995)]{f95}
596: Ferraro, F.R., Fusi Pecci, F., Testa, V., Greggio, L., Corsi, C.E., Buonanno,
597: R., Terndrup, D.M., \& Zinnecker, H., 1995, \mnras, 272, 391
598: \bibitem[Ferraro et al.(2004a)]{f04}
599: Ferraro, F. R., Origlia, L., Testa, V. \& Maraston, C., 2004, \apj, 608, 772
600: \bibitem[Ferraro et al.(2004b)]{f04b}
601: Ferraro, F. R., Beccari, G., Rood, R. T., Bellazzini, M., Sills, A., \& Sabbi, E.,
602: 2004, \apj, 603, 127
603: \bibitem[Ferraro et al.(2006)]{f06}
604: Ferraro, F. R., Mucciarelli, A., Carretta, E., \& Origlia, L., 2006, \apj, 645, L33
605: \bibitem[Freytag, Ludwig \& Steffen(1996)]{freytag}
606: Freytag, B., Ludwig, H.-G., \& Steffen, M., 1996, \aap, 313, 497
607: \bibitem[Fukunaga(1972)]{fuk}
608: Fukunaga, K., 1972, "Introduction to statististical pattern recognition", Academic Press, New York
609: \bibitem[Gallart et al.(2003)]{gallart03}
610: Gallart, C., Zoccali, M., Bertelli, G., Chiosi, C., Demarque, P.,
611: Girardi, L., Nasi, E., Woo, J.-H., \& Yi, S., 2003, \aj, 125, 742
612: \bibitem[Geisler \& Hodge(1980)]{geisler}
613: Geisler D., \& Hodge, P., 1980, \apj, 242, 73
614: \bibitem[Geisler et al.(1997)]{geisler97}
615: Geisler, D., Bica, E., Dottori, H., Claria, J. J., Piatti, A. E., \& Santos, J. F. C. Jr.,
616: 1997, \aj, 114, 1920
617: \bibitem[Girardi et al.(1995)]{gir95}
618: Girardi, L., Chiosi, C., Bertelli, G., \& Bressan, A. 1995, \aap, 298, 87
619: \bibitem[Iben et al.(1968)]{iben68}
620: Iben, I., 1968, Nature, 220, 143
621: \bibitem[Kerber, Santiago \& Brocato(2007)]{kerber}
622: Kerber, L., Santiago, B., \& Brocato, E.,2007, \aap, 462, 139
623: \bibitem[Mackey \& Gilmore(2003)]{mac03}
624: Mackey, A. D. \& Gilmore, G. F., 2003, \mnras, 338, 85
625: \bibitem[Mackey \& Gilmore(2004)]{mac04}
626: Mackey, A. D. \& Gilmore, G. F., 2004, \mnras, 352, 153
627: \bibitem[Mackey, Payne \& Gilmore(2006)]{mac06}
628: Mackey, A. D., Payne, M. J., \& Gilmore, G. F., 2006, \mnras, 369, 921
629: \bibitem[Mackey \& Broby Nielsen(2007)]{mac07}
630: Mackey, A. D., \& Broby Nielsen, P., 2007, astro-ph/0704336
631: \bibitem[Mateo(1988)]{mateo}
632: Mateo, M., 1988, \apj, 331, 261
633: \bibitem[Meylan \& Heggie(1997)]{meylan97}
634: Meylan, G., \& Heggie, D. C., 1997, \araa, 8, 1
635: \bibitem[Montegriffo et al.(1995)]{mon95}
636: Montegriffo, P., Ferraro, F. R., Fusi Pecci, F., \& Origlia, L., 1995, \mnras, 276, 739
637: \bibitem[Mould \& Aaronson(1979)]{ma79}
638: Mould, J., \& Aaronson, M., 1979, 1979, BAAS, 11, 719
639: \bibitem[Mould \& Aaronson(1982)]{ma82}
640: Mould, J., \& Aaronson, M., 1982, \apj, 263, 629
641: \bibitem[Mould et al.(1989)]{mould}
642: Mould, J., Jerome, K., Nemec, J., Jensen, J., \& Aaronson, M., 1989, \apj, 339, 84
643: \bibitem[Mucciarelli et al.(2006)]{m06}
644: Mucciarelli, A., Origlia, L., Ferraro, F. R., Maraston, C., \& Testa, V., 2006, \apj,
645: 646, 939
646: \bibitem[Mucciarelli et al.(2007)]{m07}
647: Mucciarelli, A., Ferraro, F. R., Origlia, L., \& Fusi Pecci, F., 2007, \aj, 133, 2053
648: \bibitem[Olsen et al.(1998)]{olsen98}
649: Olsen, K. A. G., Hodge, P. W., Mateo, M., Olszewski, E. W.,
650: Schommer, R. A., Suntzeff, N. B., \& Walker, A. R., 1998, \mnras, 300, 665
651: \bibitem[Origlia \& Leitherer(2000)]{origlia}
652: Origlia, L., \& Leitherer, C., 2000, \aj, 119, 2018
653: \bibitem[Persson et al.(1983)]{persson83}
654: Persson, S. E., Aaronson, M., Cohen, J. G., Frogel, J. A., \& Matthews, K.,1983, \apj, 266, 105
655: \bibitem[Salaris et al.(1993)]{salaris}
656: Salaris, M., Chieffi, A., \& Straniero, O., 1993, \apj, 414, 580
657: \bibitem[Searle, Wilkinson, \& Bagnuolo(1980)]{swb}
658: Searle, L., Wilkinson, A., \& Bagnuolo, W. G. 1980, \apj, 239, 803
659: \bibitem[Sigurdsson \& Phinney(1995)]{sig}
660: Sigurdsson, S., \& Phinney, E. S., 1995, \apjs, 99, 609
661: \bibitem[Stetson(1987)]{stet}
662: Stetson, P. B., 1987, \pasp, 99, 191
663: \bibitem[Testa et al.(1995)]{testa95}
664: Testa, V., Ferraro, F. R., Brocato, V., \& Castellani, V., 1995,
665: \mnras, 275, 454.
666: \bibitem[Testa et al.(1999)]{testa99}
667: Testa, V., Ferraro, F. R., Chieffi, A., Straniero, O., Limongi,M.,
668: \& Fusi Pecci, F., 1999, \aj, 118, 2839
669: \bibitem[Vallenari et al.(1994)]{valle94}
670: Vallenari, A., Aparicio, A., Fagotto, F., \& Chiosi, C., 1994, \aj, 284, 424
671: \bibitem[van den Bergh(1981)]{vdb81}
672: van den Bergh, S., 1981, \aaps, 46, 79
673: \end{thebibliography}
674:
675: \newpage
676:
677:
678: \begin{figure}[h]
679: \plotone{fig1.ps}
680: \caption{ACS/WFC F814W image of the LMC cluster NGC 1783, both two chips.}
681: \label{fits}
682: \end{figure}
683:
684:
685: \begin{figure}[h]
686: \plotone{fig2.ps}
687: \caption{(F555W, F555W-F814W) CMD of the LMC cluster NGC 1783,
688: obtained with ACS@HST (only stars lying into the chip containing the cluster
689: core have been plotted).}
690: \label{cmd}
691: \end{figure}
692:
693:
694: \begin{figure}[h]
695: \plotone{fig3.ps}
696: \caption{Radial (F555W ,F555W-F814W) CMD of NGC 1783 at increasing distances from the cluster
697: center. }
698: \label{radcmd}
699: \end{figure}
700:
701:
702: \begin{figure}[h]
703: \plotone{fig4.ps}
704: \caption{Completeness curves computed in two radial sub-regions of NGC 1783.
705: The black points indicate the value of the $\phi=\frac{N_{rec}}{N_{sim}}$
706: parameter calculated for each
707: 0.5 magnitude bin. }
708: \label{comp}
709: \end{figure}
710:
711: \begin{figure}[h]
712: \plotone{fig5.ps}
713: \caption{{\it Upper panel}: the map of NGC 1783 with the isodensity contours;
714: {\it central panel}:
715: the best fit ellipses to the isodensity contours; {\it lower panel}: ellipticity of
716: the best fit ellipses as a function of the semi-major axis in arcsec. The
717: horizontal dashed line indicates the mean value.}
718: \label{isoden}
719: \end{figure}
720:
721: \begin{figure}[h]
722: \plotone{fig6.ps}
723: \caption{{\it Upper panel}: observed radial density profile for the cluster NGC 1783.
724: The solid line is the best fit King model, with $r_c$=24.5'' and c=1.16.
725: The {\it horizontal dashed line}
726: indicate the background level. {\it Lower panel}:
727: the $\chi^2$ test for the observed radial density profile and best-fit King model
728: (solid line). }
729: \label{radprof}
730: \end{figure}
731:
732:
733: \begin{figure}[h]
734: \plotone{fig7.ps}
735: \caption{Best-fit theoretical PEL isochrones overplotted on the observed
736: CMD of NGC 1783. Models with different assumptions of the overshooting
737: efficiency ($\Lambda_{os}$) are used: the best fit age,
738: distance modulus and reddening (see text) for each
739: choice of $\Lambda_{os}$ are also marked.}
740: \label{isoc}
741: \end{figure}
742:
743: \begin{figure}[h]
744: \plotone{fig8.ps}
745: \caption{
746: Portion of the CMD, as zoomed onto the TO region,
747: with different models overplotted.
748: Panel (a): best-fit (dashed line, $\tau$=0.9 Gyr, $(m-M)_0$=18.45) and older (solid line,
749: $\tau$=1.2 Gyr, $(m-M)_0$=18.16) isochrones for $\Lambda_{os}$=0.00.
750: Panel (b): best-fit (dashed line, $\tau$=1.2 Gyr, $(m-M)_0$=18.45) and older (solid line,
751: $\tau$=1.4 Gyr, $(m-M)_0$=18.25) isochrones for $\Lambda_{os}$=0.10.
752: Panel (c): best-fit (dashed line, $\tau$=1.6 Gyr, $(m-M)_0$=18.45) and youngerer (solid line,
753: $\tau$=1.4 Gyr, $(m-M)_0$=18.66) isochrones for $\Lambda_{os}$=0.25.
754: }
755: \label{isoc2}
756: \end{figure}
757:
758: \begin{figure}[h]
759: \plotone{fig9.ps}
760: \caption{Histogram of the number of MS stars per $arcmin^{2}$
761: at r$>$150'' from the center of NGC 1978.}
762: \label{dec}
763: \end{figure}
764:
765: \begin{figure}[h]
766: \plotone{fig10.ps}
767: \caption{Panel {\sl (a)}: integrated LF of the
768: MS stars normalized to the number of
769: He-Clump stars: {\it black points} indicate the observed LF
770: and their size
771: correspond to their typical uncertainty.
772: The three line are the theoretical
773: LFs computed by adopting $\Lambda_{os}$=0.0 (dashed), 0.10 (dotted) and 0.25
774: (continuous). Panel {\sl (b)}:
775: same as panel {\sl (a)}, but adding
776: a 10\% binary fraction in the computation of the theoretical LFs.}
777: \label{lumfun}
778: \end{figure}
779:
780:
781: \begin{figure}[h]
782: \plotone{fig11.ps}
783: \caption{ACS@HST (F555W, F555W-F814W) CMDs for the
784: LMC cluster NGC 1783 (left panel) and NGC 1978 (right panel). The arrows
785: indicate the magnitude difference $\delta V_{SGB}^{He-Cl}$
786: between the He-Clump
787: and the flat portion of the SGB.}
788: \label{comcmd}
789: \end{figure}
790:
791:
792: \begin{figure}[h]
793: \plotone{fig12.ps}
794: \caption{Theoretical predictions for the magnitude difference between
795: the He-Clump and the flat portion of the SGB as a function of the age for three different
796: {\sl overshooting} assumptions: $\Lambda_{OS}$=0.0 (dashed line), $\Lambda_{OS}$=0.10
797: (dotted line) and $\Lambda_{OS}$=0.25 (solid line).
798: The observed values for $\delta V_{SGB}^{He-Cl}$ and the
799: inferred ages for NGC 1783 (this paper) and NGC 1978 (Paper I) are plotted as
800: black points.}
801: \label{theo}
802: \end{figure}
803:
804:
805:
806: \end{document}
807:
808: