1: %\documentstyle[10pt,emulateapj]{article}
2: %\documentstyle[12pt,aaspp4]{article}
3: %\documentclass{aastex}
4: %\documentclass[preprint2]{aastex}
5:
6: \documentclass[12pt,preprint]{aastex}
7: %\documentclass[manuscript]{aastex}
8: %\singlespace
9:
10: % \usepackage{emulateapj5}
11:
12: \def\p/{\mbox{$^1$}}
13: \def\pp/{\mbox{$^2$}}
14: \def\ppp/{\mbox{$^3$}}
15: \def\pppp/{\mbox{$^4$}}
16: \def\m/{\mbox{$^{-1}$}}
17: \def\mm/{\mbox{$^{-2}$}}
18: \def\mmm/{\mbox{$^{-3}$}}
19: \def\mmmm/{\mbox{$^{-4}$}}
20: \def\Ms/{\mbox{M$_\odot$}}
21: \def\ebv{\mbox{$E(4405-5495)$}}
22: \def\rv{\mbox{$R_{5495}$}}
23: \def\hii{\ion{H}{2}}
24: \newcommand{\sci}[2]{\mbox{$#1\cdot 10^{#2}$}}
25: \newcommand{\scipm}[3]{\mbox{$(#1\pm #2)\cdot 10^{#3}$}}
26: %stopzone
27:
28: \shorttitle{Progenitor of SN 2004dj}
29: \shortauthors{Ma\'{\i}z-Apell\'aniz et al.}
30:
31: \slugcomment{Submitted to Astrophysical Journal Letters}
32:
33: \begin{document}
34:
35: \title{The Progenitor of the Type II-P SN 2004dj in NGC 2403}
36:
37: \author{J. Ma\'{\i}z-Apell\'aniz\altaffilmark{1,2},
38: Howard E. Bond\altaffilmark{3}, and M. H. Siegel\altaffilmark{4}}
39: \affil{Space Telescope Science Institute\altaffilmark{5}, 3700 San Martin
40: Drive, Baltimore, MD 21218, U.S.A.}
41:
42: \author{Y. Lipkin, D. Maoz, E. O. Ofek, and D. Poznanski}
43: \affil{School of Physics and Astronomy, Tel-Aviv University, Tel-Aviv 69978,
44: Israel}
45:
46: %\email{jmaiz@stsci.edu, bond@stsci.edu, siegel@astro.as.utexas.edu,
47: %yiftah@wise1.tau.ac.il, dani@wise1.tau.ac.il, eran@wise1.tau.ac.il,
48: %dovip@wise1.tau.ac.il}
49:
50: % ---------------affiliations of the science group -----------------------
51:
52: \altaffiltext{1}{Affiliated with the Space Telescope Division of the European
53: Space Agency, ESTEC, Noordwijk, Netherlands.}
54: \altaffiltext{2}{e-mail contact: {\tt jmaiz@stsci.edu}.}
55: \altaffiltext{3}{Visiting Astronomer, Kitt Peak National Observatory, National
56: Optical Astronomy Observatory, which is operated by the Association of
57: Universities for Research in Astronomy, Inc., under cooperative agreement with
58: the National Science Foundation.}
59: \altaffiltext{4}{Current address: Department of Astronomy, University of Texas,
60: Austin, TX 78712.}
61: \altaffiltext{5}{The Space Telescope Science Institute is operated by the
62: Association of Universities for Research in Astronomy, Inc., under NASA
63: contract No. NAS5-26555.}
64:
65: \begin{abstract}
66:
67: The Type II-P supernova 2004dj in the nearby galaxy NGC~2403 occurred at a
68: position coincident with object 96 in the list of luminous stars and clusters
69: in this galaxy published by Sandage in 1984. The coincidence is established
70: definitively through astrometric registration of our ground-based archival
71: images of NGC~2403 with our recent images showing the SN\null. The archival
72: images show that Sandage~96 is slightly resolved from the ground. Pre-outburst
73: blue spectrograms obtained by Humphreys and Aaronson reveal that
74: Sandage~96 has a composite spectrum, dominated in the blue region by A- and
75: B-type stars, while infrared photometry shows that Sandage~96 also contains red
76: supergiants. These results demonstrate that Sandage~96 is a young compact
77: cluster. We have studied the stellar population of Sandage~96, using published
78: photometric measurements combined with a chi-square-fitting code. We derive a
79: cluster age of 13.6~Myr, a reddening of $\ebv=0.172$, and a total stellar mass
80: of 24,000~$M_\odot$\null. For this cluster age, the SN progenitor had a
81: main-sequence mass of $15\,M_\odot$. Post-outburst photometry of Sandage~96 may
82: establish whether the progenitor was a red or blue supergiant.
83:
84: \end{abstract}
85:
86: \keywords{galaxies: individual (NGC 2403) --- galaxies: star clusters ---
87: supernovae: general --- supernovae: individual (SN 2004dj)}
88:
89: \section{Introduction}
90:
91: Type II supernovae (SNe) are believed to arise from core collapses of evolved
92: massive stars, yet there is surprisingly little direct evidence in
93: the form of pre-outburst observations of progenitor stars to support this
94: belief. To date, the best-observed SN progenitor is the blue supergiant
95: precursor of SN~1987A in the Large Magellanic Cloud, but the pre-outburst data
96: are limited to photographic and photoelectric photometry and a photographic
97: objective-prism spectrum \citep[see][]{Walbetal89}. More recently, the
98: progenitors of (or binary companions of) two further Type II SNe have
99: been identified on archival images from the {\it Hubble Space Telescope\/}
100: ({\it HST\/}) and ground-based telescopes \citep{Maunetal04,Smaretal04}.
101: See \citet{Smaretal03} for further discussion of the importance of efforts
102: to identify SN progenitors.
103:
104: The bright SN 2004dj was discovered in the nearby spiral galaxy NGC 2403 on
105: 2004 July 31 by K.~Itagaki \citep[see][]{Naka04}, and it was quickly classified
106: spectroscopically as a normal Type II-P (plateau) core-collapse SN caught about
107: 3 weeks after the explosion \citep{Pataetal04}. NGC 2403 is a member of the
108: M81 group, with a distance of about 3.3~Mpc \citep{Karaetal04}; SN~2004dj is
109: thus the nearest SN to be discovered since SN~1993J in M81, and it is destined
110: to be a well-observed SN~II-P event. Type II-P SNe have been reviewed by
111: \citet{Leonetal02} and \citet{Hamu03}; Hamuy's Table~1, which lists the
112: properties of 24 well-documented SN~II-P events, shows that SN~2004dj is the
113: nearest Type~II-P SN yet discovered (apart from the peculiar Type~II-P
114: SN~1987A)\null.
115:
116: In this {\it Letter}, we use recent images of the SN along with archival images
117: to show that SN~2004dj coincides with a previously cataloged object in
118: NGC~2403, which proves to be a slightly resolved young cluster that contains
119: dozens of massive stars. Well-calibrated photometry and digital spectrograms
120: of this cluster exist in the literature. Using models of starburst populations,
121: we are able to estimate the age, mass, and other properties of this compact
122: cluster, and to set limits on the nature of the progenitor star.
123:
124: \section{Identification of the Progenitor}
125:
126: \subsection{Sandage 96 in NGC 2403}
127:
128: \citet{Sand84} conducted a photographic survey of the brightest blue and red
129: supergiants in NGC~2403. Based on the reported position of SN~2004dj, several
130: authors \citep[see][]{Yamaetal04} have pointed out the close coincidence
131: of the SN with a luminous blue object in Sandage's list, cataloged as no.~96.
132: This object is designated ``Sandage~96'' hereafter.
133:
134: Sandage 96 would be one of the brightest blue supergiants in NGC~2403, if it
135: were a single star. However, \citet{Sand84} annotated it as a possible
136: cluster. \citet{Hump80} obtained a blue spectrum of this object and classified
137: it as B5:~I:, but suggested a composite nature. An improved blue spectrum was
138: obtained by \citet{HumpAaro87}, and is reproduced in their Figure~1. They
139: report that the integrated spectrum resembles that of an early A-type star, but
140: that the hydrogen Balmer lines are too broad to be due solely to an A
141: supergiant, and the presence of \ion{He}{1} lines shows that there are
142: also B-type stars present in the spectrum. Humphreys \& Aaronson also did not
143: detect H$\alpha$ emission. They concluded that Sandage~96 is a compact cluster,
144: without an accompanying \ion{H}{2} region.
145:
146: Sandage~96 was included as object n2403-3866 in a list of young massive
147: clusters in nearby galaxies by \citet{Lars99}, who measured
148: Johnson-Kron-Cousins $UBVI$ photometry. The object is also present in the 2MASS
149: infrared point-source catalog \citep{Skruetal97}, from which we obtained its
150: $JHK_{\rm s}$ photometry. The optical and infrared photometric data from these
151: sources are given in Table~\ref{photometry}, where we have removed the
152: reddening correction applied by Larsen to his measurements. The data definitely
153: do not correspond to any single star, since the color is blue at short
154: wavelengths but an infrared excess begins to set in around the $I$ band and is
155: pronounced in the infrared.
156:
157: \subsection{Archival and New Images}
158:
159: The site of SN 2004dj was imaged (serendipitously) by H.E.B. on three nights in
160: 1999 January with the Mosaic CCD camera on the Mayall 4-m telescope at Kitt
161: Peak National Observatory (KPNO). All of the observations were made with
162: standard Johnson-Kron-Cousins $B$, $V$, and $I$ filters, along with a
163: Thuan-Gunn $u$ filter.
164:
165: % The frames were taken under good seeing conditions.
166:
167: % At the 0.9-m, standard stars from \citet{Land92} and \citet{SiegBond04} were
168: % observed on the same nights, allowing calibration of the images. For
169: % Sandage~96, we find $V=17.93\pm0.05$, $B-V=0.17\pm0.02$, $V-I=0.81\pm0.05$, and
170: % $u-B=0.41\pm0.06$ (where the $u-B$ index is on the system described by
171: % \citealt{Bond04}). Our $BVI$ photometry generally agrees well with that
172: % reported by \citet{Lars99}, as given in our Table~\ref{photometry}. The $u-B$
173: % index (after correction for a plausible amount of reddening) is characteristic
174: % of a late B type, while the $B-V$ color suggests an A type, and the $V-I$ color
175: % is that of an F type. The $uBVI$ photometry thus also suggests a strongly
176: % composite nature.
177:
178: Following the discovery announcement of SN~2004dj, CCD images of the SN were
179: obtained by Y.L., D.M., E.O.O., and D.P., using the 1.0-m telescope of the Wise
180: Observatory and $UBVRI$ filters. Figure~1 (left) shows the field of the SN and
181: a few neighboring bright field stars, as imaged in the $V$ band with the Wise
182: telescope on 2004 August~4. In Figure~1 (right) we show the same field, as
183: extracted from an $I$-band KPNO Mosaic frame taken on 1999 January~19. In
184: order to verify the association of the SN with Sandage~96 which is strongly
185: suggested by Fig.~1, we combined four of the best Wise Observatory CCD frames
186: of SN 2004dj to produce a fairly deep image. We then registered this frame
187: with the Kitt Peak 4-m frame shown in Fig.~1, using 8 nearby field stars to
188: determine the geometric transformation of the Wise frame onto the Kitt Peak
189: image. The rms of the astrometric fit is only $0\farcs03$ in each coordinate,
190: in spite of the $2\farcs6$ seeing of the Wise image and the fact that several
191: of the reference stars are saturated in the KPNO image. We find that SN~2004dj
192: coincides with Sandage~96 to within $0\farcs07$ ($0\farcs04$ in RA, $0\farcs06$
193: in Dec).
194:
195: Sandage~96 is slightly non-stellar on our archival frames. In our $I$ frame
196: with the best seeing, $0\farcs8$, Sandage~96 has a FWHM of $1\farcs0$.
197: S.~Smartt (2004, private communication) has informed us that archival images
198: taken with the 8.2-m Subaru telescope in 0\farcs44 seeing through an H$\alpha$
199: filter (which just measures continuum flux as there is no line emission) show
200: a FWHM for Sandage~96 of 0\farcs6.
201:
202: It should be noted that, at the distance of NGC~2403, $1\arcsec$ corresponds
203: to a linear scale of 16~pc and that 4~pc is a typical size for a compact young
204: stellar cluster (see, e.g. \citealt{Maiz01b}). There is thus no doubt that the
205: SN lies well within the bounds of the cluster Sandage~96, and that it must have
206: arisen from a star belonging to the cluster.\footnote{After this letter was
207: submitted, {\it HST\/} observations of SN~2004dj were obtained on 2004 August
208: 16 and 17. Based on long-exposure ACS/WFC images, \citet{FiliLi04} report (and
209: we confirm independently) that, although the SN is heavily saturated in these
210: frames, it definitely took place at a position coincident with
211: Sandage~96. We have analyzed higher-resolution ACS/HRC images that were also
212: obtained and find that: (a)~in the short-exposure (non-saturated) images only a
213: very bright point-source (the SN itself) is present, which is so bright that it
214: makes detection of any other stars in the cluster extremely difficult; and
215: (b)~in the NUV-optical objective-prism exposures the SN is easily detected, as
216: well as three nearby (within a few arcseconds) stars also detected in the
217: ground-based KPNO image. All of the {\it HST\/} images confirm the coincidence
218: between SN 2004dj and Sandage~96, but it will take additional observations,
219: after the SN has subsided, to determine the color-magnitude diagram and other
220: properties of the cluster.}
221:
222: % These observations clearly establish that Sandage~96 is a compact cluster in
223: % NGC~2403, and that SN~2004dj arose from one of the stars within the cluster.
224:
225: \section{The Stellar Population of Sandage 96}
226:
227: \subsection{Mathematical Approach}
228:
229: In order to determine the nature of the stellar population within the compact
230: cluster and thus constrain the main-sequence mass and other properties of the
231: SN~2004dj progenitor, we used CHORIZOS \citep{Maiz04c}, a
232: chi-square-minimization code that finds which members of a family of spectral
233: energy distributions (SEDs) are compatible with the observed integrated colors
234: of a stellar population. CHORIZOS allows the user to select different input SED
235: families and extinction laws, and to place statistical constraints on the
236: fitted parameters.
237:
238: Although the above discussion shows that Sandage~96 is a cluster, for
239: completeness we tested whether this object could have been a single luminous
240: supergiant or had to be a compact cluster, by using two different SED families,
241: one consisting of single stars, and the other of cluster populations. For the
242: stellar models we selected \citet{Kuru04} atmospheres with low gravities, solar
243: metallicity, and effective temperatures between 3,500 K and 50,000 K\null. For
244: the cluster models, we selected Starburst99 \citep{Leitetal99} model
245: populations with solar metallicity, a Salpeter IMF, an upper-mass cutoff of
246: $100 \, M_\odot$, and ages between $10^6$ and $10^{10}$~yr. The starburst models
247: assume that all of the stars in the population were created at the same time.
248: The choice of solar metallicities for the models is justified by the
249: galactocentric distance of Sandage 96 \citep{Fieretal86}.
250:
251: CHORIZOS was executed using both SED families, selecting a \citet{Cardetal89}
252: interstellar extinction law with $\rv = 3.1$, and two free parameters: the
253: extinction, \ebv, and either the effective temperature (for the stellar models)
254: or log(age) (for the cluster models).\footnote{\ebv\ and $R_{5495}$ are the
255: monochromatic equivalents to $E(B-V)$ and $R_V$, respectively. Here 4405 and
256: 5495 are the assumed central wavelengths (in \AA) of the $B$ and $V$ filters,
257: respectively. CHORIZOS uses monochromatic quantities because $E(B-V)$ and $R_V$
258: depend not only on the amount and type of dust but also on the SEDs.} Since we
259: are fitting two parameters and using six colors (derived from seven
260: magnitudes), the problem has 4 degrees of freedom.
261:
262: \subsection{Results of the Population Fitting}
263:
264: Our results from CHORIZOS decisively eliminate the possibility that Sandage~96
265: is a single luminous star. The best fit for the Kurucz models is
266: for a highly reddened O~star, but its reduced $\chi^2$ is
267: 43, indicating the extremely poor quality of the fit.
268:
269: However, cluster models provide an excellent fit, with a best
270: reduced $\chi^2$ of 0.28 (which, if anything, suggests that photometric
271: uncertainties may have been overestimated). The likelihood map produced by
272: CHORIZOS (Fig.~\ref{likelihood}) shows two peaks in the log(age)-\ebv\ plane,
273: indicating the existence of two solutions compatible with the available
274: photometry. Properties for both solutions are shown in
275: Table~\ref{results}. The ``young'' solution (age of 13.6~Myr) is the one that
276: has the highest likelihood. The ``old'' solution (age around 29~Myr) is less
277: likely, but its validity cannot be immediately rejected at the 10\% level,
278: since its reduced $\chi^2$ is 1.74. The SED for the ``young''
279: solution is shown in Fig.~\ref{spectra}.
280:
281: Fig.~2 shows that the fit to the ``young'' solution is excellent. For the
282: ``old'' solution, the largest contribution to $\chi^2$ is from the
283: $U$ band, because the Balmer jump of the SED is large compared
284: to
285: the observed photometry (and to the spectrum shown by \citet{HumpAaro87}.
286: Therefore, we strongly prefer the ``young'' solution.
287:
288: This solution yields a reddening of $\ebv=0.172\pm0.022$,
289: which
290: is in very good agreement with
291: the $E(B-V) = 0.18$ obtained by \citet{Pataetal04} from the \ion{Na}{1} D
292: equivalent
293: width in the SN spectrum.
294:
295: The turnoff mass for a cluster at an age of 13.6~Myr is $15\,M_\odot$, and the
296: corresponding main-sequence spectral type is B1~V.
297:
298: \section{Discussion}
299:
300: The age of 13.6~Myr derived above is also compatible with other
301: secondary evidence. The blue spectrum of a 13.6 Myr old cluster is dominated by
302: blue giants and supergiants, so its classification as composite B5: I: by
303: \citet{Hump80} agrees with our expectation. Furthermore, NGC~2403 is a galaxy
304: that has undergone recent intense episodes of star formation that have produced
305: several massive young clusters and OB associations
306: \citep{Drisetal99b,Maiz01b}. Some of those clusters are only a few Myr old and
307: show strong H$\alpha$ emission; however, as noted above, Sandage~96 is not
308: detected in H$\alpha$ \citep[see also][]{Sivaetal90}. This is what is expected
309: for a 13.6 Myr old cluster, since such an object would have had enough time to
310: disperse its parent molecular cloud by stellar winds and supernova explosions,
311: and should have no O stars left [apart from, possibly, O stars formed as a
312: result of mass transfer in binaries \citep{Cerv98}].
313:
314: From the known distance and measured $V$ magnitude, we can determine the total
315: mass of the stellar population\footnote{The mass estimate assumes a Salpeter
316: IMF between $1 M_\odot$ and $100 M_\odot$.}. The result,
317: $\approx$24,000$\,M_\odot$, makes Sandage~96 intermediate between the most
318: massive young clusters in NGC~2403 and typical Galactic open clusters, with a
319: mass similar to those of some of the ``rich'' young clusters in the LMC\null.
320: Therefore, Sandage~96 is likely not massive enough to survive for a Hubble time
321: and become a future globular cluster; rather it is destined to dissolve into
322: the general field of NGC~2403 \citep{FallZhan01}.
323:
324: The uncertainties quoted in Table~\ref{results} should be taken with caution,
325: since they do not include external error sources, such as uncertainties in the
326: distance, stellar models, and other assumptions, including the assumed
327: instantaneous starburst episode. In particular, we should note that current
328: stellar evolutionary models have problems in producing the right numbers and
329: types of red supergiants in the 6-30 Myr age interval
330: \citep{LangMaed95,MassOlse03}. Therefore, it is possible that the actual age
331: for Sandage~96 may differ from our value by 1 or 2~Myr. Unless the cluster is
332: very compact, future {\it HST\/} imaging should provide a color-magnitude
333: diagram that will clarify these issues.
334:
335: Another point is that our starburst models assume an infinite number of stars,
336: whereas the actual number of bright stars in Sandage~96 must be relatively low,
337: leading to a potential problem of stochastic sampling. However, as we show in
338: Table~\ref{results}, the predicted numbers of red supergiants in Sandage~96
339: (the main contributors to the NIR SED) and of blue giants and supergiants (the
340: main contributors to the blue-visible spectrum) are above the critical value of
341: 10 \citep{CervVall03} below which sampling effects start to become significant.
342:
343: Supernovae of Type II-P are expected to have red supergiant progenitors with
344: initial masses lower than those of Type II-L and Ib/c \citep{Hegeetal03}. The
345: turnoff mass we find for Sandage~96, $15\,M_\odot$, is consistent with that of
346: other Type II-P SNe \citep{Smaretal03}. However, it should be pointed out that
347: the heretofore best-studied SN progenitor of a core-collapse supernova, that of
348: SN~1987A, turned out to be a blue supergiant. As seen in Table~\ref{results},
349: our models predict that $\approx$12 red and 2-3 blue supergiants should have
350: been present in Sandage~96 prior to the explosion of SN~2004dj. Given that
351: red supergiants are the dominant sources of the NIR flux in a 13.6~Myr old
352: cluster, and that blue giants and supergiants dominate in the $U$ and $B$
353: bands, we can make a prediction that will test the nature of the progenitor.
354: Once the SN has faded away, if the star that exploded was a red supergiant,
355: then the NIR integrated magnitudes of the cluster should have dimmed by about
356: 0.08~mag, while the blue magnitudes will be essentially unchanged. If, on the
357: other hand, the progenitor was a blue supergiant, the $UBV$ photometry should
358: be dimmer by $\approx$0.04~mag. However, it will take several years before the
359: SN has faded enough to make a negligible contribution to the total flux of
360: Sandage~96 \citep{Pozzetal04,Zhanetal04}.
361:
362: We conclude that SN 2004dj took place in Sandage~96, an $\approx$14 Myr old
363: compact cluster of intermediate mass. Based on its SN type of II-P and the
364: $15\,M_\odot$ turnoff mass of the cluster, the progenitor is likely to have
365: been a red supergiant, of which $12\pm4 $ existed in Sandage~96 prior to the
366: explosion. Although a blue-supergiant progenitor cannot be ruled out completely
367: at this stage, our population models suggest that a comparison of pre- and
368: post-outburst photometry of Sandage~96 may determine whether the exploded
369: progenitor was blue or red.
370:
371: \begin{acknowledgements}
372:
373: We would like to thank Nolan Walborn for helpful discussions; Leonardo \'Ubeda
374: for help with evolutionary tracks; and the referee, Stephen Smartt, for very
375: helpful suggestions and for communicating results in advance of publication.
376: H.E.B. acknowledges support from the NASA
377: UV, Visible, and Gravitational Astrophysics Research and
378: Analysis Program through grant NAG5-3912. D.M. acknowledges support by the
379: Israel Science Foundation---the Jack Adler Foundation for Space Research, Grant
380: 63/01-1.
381:
382: \end{acknowledgements}
383:
384: \bibliographystyle{apj}
385: \bibliography{general}
386:
387: \begin{deluxetable}{cr}
388: \tablecaption{Published Photometry}
389: \tablewidth{0pt}
390: \tablehead{\colhead{Quantity} & \colhead{Value}}
391: \startdata
392: $V$ & 18.05\phn\phn$\pm$\phn0.03\phn \\
393: $U-B$ & $-$0.48\phn\phn$\pm$\phn0.06\phn \\
394: $B-V$ & 0.30\phn\phn$\pm$\phn0.03\phn \\
395: $V-I$ & 0.87\phn\phn$\pm$\phn0.03\phn \\
396: $J$ & 16.193\phn$\pm$\phn0.081 \\
397: $H$ & 15.539\phn$\pm$\phn0.106 \\
398: $K_{\rm s}$ & 15.416\phn$\pm$\phn0.191 \\
399: \enddata
400: \label{photometry}
401: \end{deluxetable}
402:
403:
404:
405: \begin{deluxetable}{lll}
406: \tablecaption{Results from CHORIZOS Fitting for Sandage 96.}
407: \tablewidth{0pt}
408: \tablehead{ & \colhead{Young model} & \colhead{Old model}}
409: \startdata
410: Age (Myr) & \phs13.6\phn\phn\phn$\pm$\phn0.2 & \phs28.8\phn\phn\phn$\pm$\phn2.2 \\
411: \ebv\ (mag) & \phs\phn0.172\phn$\pm$\phn0.022 & \phs\phn0.277\phn$\pm$\phn0.021 \\
412: $M_V$ & $-$10.08\phn\phn$\pm$\phn0.09 & $-$10.39\phn\phn$\pm$\phn0.07 \\
413: Mass ($10^3\,M_\odot$)\tablenotemark{a} & \phs24.0\phn\phn\phn$\pm$\phn1.1 & \phs57.5\phn\phn\phn$\pm$\phn2.6 \\
414: $\chi^2$ per degree of freedom & \phs\phn0.28 & \phs\phn1.74 \\
415: K+M stars, types I and II\tablenotemark{b} & \phs12 & \phs50 \\
416: B stars, types I and II\tablenotemark{b} & \phs\phn2.5 & \phs17 \\
417: B stars, type III\tablenotemark{b} & \phs24 & \phs89 \\
418: \enddata
419: \tablenotetext{a}{Assuming a Salpeter IMF between $1\,M_\odot$ and
420: $100\,M_\odot$.}
421: \tablenotetext{b}{Predicted number of stars, from Starburst99.}
422: \label{results}
423: \end{deluxetable}
424:
425:
426: \begin{figure}
427: \begin{center}
428: \centerline{\includegraphics*[width=0.48\linewidth]{wise.eps} \
429: \includegraphics*[width=0.48\linewidth]{kpno.eps}}
430: \end{center}
431: \figcaption{CCD images of the site of SN 2004dj taken on 2004 August 4
432: (left; Wise Observatory 1.0-m, $V$ band, 60-s exposure) and 1999 January 19
433: (right; Kitt Peak 4-m, $I$ band, 600-s exposure). North is at the top
434: and east is on the left, and the images are $96\arcsec$ wide. The supernova
435: coincides with Sandage~96 to within $0\farcs07$.}
436: \end{figure}
437:
438: \clearpage
439:
440: \begin{figure}
441: % \centerline{\includegraphics*[width=0.36\linewidth, bb=28 28 566 514]{kurucz00001_lowg_spect.ps}}
442: \centerline{\includegraphics*[width=\linewidth, bb=28 28 566 514]{sb9900001_spect.ps}}
443: \caption{SED for the best fit (13.6 Myr solution) to the optical-NIR photometry of
444: Sandage~96, based on Starburst99 models of integrated stellar populations.
445: The photometry itself is shown by the symbols with error bars (vertical ones
446: for uncertainties and horizontal ones for the approximate wavelength coverage
447: of each filter). Star symbols indicate the calculated magnitude of the model
448: SED for each filter.}
449: \label{spectra}
450: \end{figure}
451:
452: \begin{figure}
453: \centerline{\includegraphics*[width=0.8\linewidth, bb=28 28 566
454: 538]{sb9900001_b.ps}}
455: \caption{Likelihood contour plot in reddening vs.\ age for the CHORIZOS fit
456: using Starburst99 models. Age is expressed in years. The favored solution has a
457: cluster age of 13.6 Myr and a reddening of 0.172~mag.}
458: \label{likelihood}
459: \end{figure}
460:
461:
462: \end{document}
463:
464: