1: %\documentclass[12pt,preprint]{aastex}
2: %\documentclass[11pt,manuscript]{aastex}
3: \documentclass[apj]{emulateapj}
4:
5: %% Running heads
6: \shorttitle{Stellar Content \& Recent SFH of IC1613}
7: \shortauthors{Bernard et al.}
8:
9:
10: \begin{document}
11:
12: %\title{Stellar Content and Recent Star Formation History of \\ the Local Group
13: % Dwarf Irregular Galaxy IC1613\footnote{Based on observations made with the
14: % Isaac Newton Telescope operated on the island of La Palma by the Isaac
15: % Newton Group in the Spanish Observatorio del Roque de los Muchachos of the
16: % Instituto de Astrof\'{i}sica de Canarias}}
17: \title{Stellar Content and Recent Star Formation History of \\ the Local Group
18: Dwarf Irregular Galaxy IC1613\footnotemark[1]}
19:
20: \author{Edouard J. Bernard, Antonio Aparicio\altaffilmark{2}, Carme Gallart,
21: Carmen P. Padilla-Torres, Maurizio Panniello$\dagger$}
22: \affil{Instituto de Astrof\'{i}sica de Canarias, E-38205 La Laguna,
23: Tenerife, Spain}
24: \email{ebernard@iac.es, antapaj@iac.es, carme@iac.es, cppt@iac.es.}
25:
26: \footnotetext[1]{Based on observations made with the Isaac Newton Telescope
27: operated on the island of La Palma by the Isaac Newton Group in the Spanish
28: Observatorio del Roque de los Muchachos of the Instituto de Astrof\'{i}sica
29: de Canarias.}
30: \altaffiltext{2}{Departamento de Astrof\'{i}sica, Universidad de La Laguna,
31: E-38200 La Laguna, Tenerife, Spain.}
32: \altaffiltext{$\dagger$}{Deceased.}
33:
34: \begin{abstract}
35:
36: We present resolved-star $VI$ photometry of the Local Group dwarf
37: irregular galaxy IC1613 reaching I\,$\sim$\,23.5, obtained with the
38: wide-field camera at the 2.5\,m Isaac Newton Telescope.
39: A fit to the stellar density distribution shows an exponential profile
40: of scale length $2\farcm9 \pm 0.1$ and gives a central surface
41: brightness $\mu_{V,0} = 22.7 \pm 0.6$.
42: The significant number of red giant branch (RGB) stars present in the
43: outer part of our images ($r > 16\farcm5$) indicates that the galaxy
44: is actually more extended than previously estimated.
45: A comparison of the color-magnitude diagrams (CMDs) as a function of
46: galactocentric distance shows a clear gradient in the age of its
47: population, the scale length increasing with age, while we find no
48: evidence of a metallicity gradient from the width of the RGB.
49: We present quantitative results of the recent star formation history
50: from a synthetic CMD analysis using IAC-STAR. We find a mean star formation
51: rate of $(1.6 \pm 0.8) \times 10^{-3} $~M$_\Sun~$yr$^{-1}~$kpc$^{-2}$ in
52: the central $r \la 2\farcm5$ for the last 300~Myr.
53:
54: \end{abstract}
55:
56:
57: \keywords{galaxies: dwarf ---
58: galaxies: individual (IC1613) ---
59: galaxies: irregular ---
60: galaxies: stellar content ---
61: galaxies: structure ---
62: Local Group}
63:
64:
65: \section{Introduction}
66:
67: \setcounter{footnote}{2}
68:
69: IC1613 is a typical dwarf irregular galaxy concerning its
70: luminosity, metallicity and star formation rate (SFR). In fact, it
71: serves as the prototype for the DDO type Ir~V.
72: It is a low surface-brightness galaxy with a moderate luminosity
73: \citep[$M_V=-15.3$,][]{ber00} located $730 \pm 20$ kpc from our
74: Sun. This distance corresponds to a distance modulus
75: $(m-M)_0=24.31 \pm 0.06$ \citep{dol01} and a scale of 3.54 pc per
76: arcsecond. The recent calculation of \citet{pie06} from near-IR
77: photometry of cepheids gives $(m-M)_0=24.291 \pm 0.035$, for which
78: the authors claim a total uncertainty of less than 3\%.
79: While the spatial extent estimate of \citet{abl72} gives an optical
80: size of $16\arcmin \times 20\arcmin$, the observation of carbon stars
81: up to 15$\arcmin$ from the center by \citet{alb00} indicate that it
82: is virtually twice that size.
83: Its high galactic latitude, in the southern hemisphere, confers it
84: low extinction and color excess. Here we adopt a reddening of
85: $E(B-V)=0.02 \pm 0.02$ from \citet{col99}.
86:
87: The H\,\textsc{II} region metallicity of IC1613 was measured by
88: \citet[see \citealt{ski89}]{tal80}
89: to be 12\,+\,log(O/H)~$=7.86$ from spectrophotometric observations
90: of the [O\,\textsc{III}]$\lambda4363$ line, while \citet{lee03}
91: found 12\,+\,log(O/H)~$=7.62$.
92: This corresponds to [Fe/H]~$=-0.8$ and [Fe/H]~$=-1.07$ dex
93: respectively \citep[hereinafter S03]{ski03}.
94: The mean color $(V-I)_{-3.5}$ of the RGB at $M_I$~=~-3.5 gives
95: [Fe/H] = -1.3 for the old and intermediate-age population \citep{lee93}.
96: This low overall metallicity and the high gas content \citep{hof96}
97: suggest a primitive state in its evolution.
98:
99: %%% Defines the aliases to use 'S03' instead of 'Skillman et al. 2003'
100: \defcitealias{ski03}{S03}
101: \defcitealias{col99}{Cole et al. 1999}
102: \defcitealias{dol01}{Dolphin et al. 2001}
103:
104: The star formation history (SFH) has been studied quite extensively by
105: \citetalias{ski03} for an HST/WFPC2 field located $7\farcm4$ kpc southwest
106: from the center. Their conclusion is a relatively constant SFR over a long
107: period, with the oldest population being more than 10~Gyr old.
108: They also summarize the results about structure and stellar content of
109: the whole galaxy from the literature. More recently, \citet{bor04}
110: analyzed 60 OB associations, apparently correlated with H\,\textsc{II}
111: regions studied kinematically by \citet[see also \citealt{sil06}]{loz03}.
112: \citet{mag05} announced the detection of two candidate
113: planetary nebulae.
114:
115: In this paper we describe a wide-field survey of IC1613: an overview of
116: the observations and data reduction is presented in section
117: \ref{datared}, and the resulting color-magnitude diagram (CMD) is
118: described in \S \ref{descCMD}.
119: In \S \ref{morpho} we examine the extent and morphology of the galaxy,
120: as well as the spatial structure of the different populations in the CMDs.
121: An analysis of the recent SFH at different galactocentric radii is given
122: in \S \ref{recentstarfo}.
123: Finally, in \S \ref{concl} we summarize the results and present
124: our conclusions.
125:
126:
127: \section{Observations \& Data Reduction}\label{datared}
128:
129: Observations of IC1613 in Harris $V$ and Sloan Gunn $i'$ were conducted on
130: 5 nights between November 1999 and September 2000 using the Wide Field
131: Camera (WFC) at the 2.5 m Isaac Newton Telescope (INT) of the
132: Observatorio del Roque de los Muchachos.
133: The WFC is a mosaic camera made up of four $2048 \times 4096$ CCDs, with
134: a pixel size of $0.33 \arcsec$. The total field of view is about $34
135: \arcmin \times 34 \arcmin$, covering most of the galaxy. The total
136: integration times in $V$ and $i'$ were 3660 and 1830 seconds, respectively.
137: A detailed observing log is presented in Table~\ref{obs}.
138:
139: \begin{deluxetable}{ccccc}
140: \tablewidth{0pt}
141: \tablecaption{Journal of Observations. \label{obs}}
142: \tablehead{
143: \colhead{UTC Date} & \colhead{Time (UT)} & \colhead{Filter} &
144: \colhead{Exp. Time (s)} & \colhead{Air Mass}}
145: \startdata
146: 1999 Nov 06 & 22:23 & V & 60 & 1.14 \\
147: 1999 Nov 06 & 22:32 & V & 1200 & 1.13 \\
148: 1999 Nov 06 & 22:53 & V & 1200 & 1.12 \\
149: 1999 Nov 06 & 23:14 & V & 1200 & 1.12 \\
150: 2000 Aug 10 & 04:38 & I & 30 & 1.12 \\
151: 2000 Aug 10 & 04:41 & I & 600 & 1.12 \\
152: 2000 Aug 10 & 04:54 & I & 600 & 1.12 \\
153: 2000 Aug 10 & 05:05 & I & 600 & 1.12 \\
154: \enddata
155: \end{deluxetable}
156:
157: Overscan trimming, bias subtraction and flat-field corrections were
158: performed using the standard routines in IRAF\footnote[3]{IRAF is
159: distributed by the National Optical Astronomy Observatory,
160: which is operated by the Association of Universities for Research
161: in Astronomy, Inc., under cooperative agreement with the National
162: Science Foundation.}. The $i'$ images have also
163: been corrected for fringing effects. The DAOPHOT-II/ALLSTAR and ALLFRAME
164: programs \citep{ste87,ste94} were then used to obtain the instrumental
165: photometry of the resolved stars from the four individual images in each
166: band and for each chip. The $\sim$200 stars used to model the
167: point-spread functions (PSFs) were carefully selected to cover the whole
168: field of view and sample the spatial variations of the PSF. The input
169: list of stars for ALLFRAME was created with DAOMASTER
170: from the ALLSTAR photometry files of the individual images. This list
171: contains all the stars that were detected on at least one image.
172: The stars with good photometry were selected among the detected objects
173: using ALLFRAME's fitting parameters $\sigma$, $\chi ^2$ and $SHARP$. Only
174: those $\sim$30000 objects with very good photometry, i.e.,
175: with $\sigma \leq 0.15$, $\chi ^2 \leq 1.1$
176: and $-0.3 \leq SHARP \leq 0.3$, were kept.
177:
178:
179: \begin{figure}[b]
180: \epsscale{0.8} % preprint
181: \epsscale{1.1} % emulateapj
182: \plotone{f1.eps}
183: \figcaption{Color-magnitude diagram of IC1613. All the stars described
184: in section \ref{datared} are plotted.
185: \label{CMD}}
186: \end{figure}
187:
188:
189: Our photometry of chips 1, 3 and 4 of the WFC\footnote[4]{See the WFC
190: User Notes at
191: http://www.ing.iac.es/Astronomy/instruments/wfc/wfc\_notes\_apr98.html.}
192: was calibrated to standard magnitudes using OGLE's photometry of the same
193: field, kindly provided by Dr.~A. Udalsky (private communication).
194: Several hundreds of stars were used for each chip, giving dispersions
195: of the fits at the centers of mass of the point distributions of about
196: 0.001.
197: Calibration of chip 2, which was outside OGLE's field, was more laborious.
198: In $V$, one standard star field observed during the IC1613 INT run
199: was used to determine the transformations between chip 2 and chips 1 and 4.
200: The dispersion of these transformations is $\sim$0.005.
201: Then the transformation for the latter chips based on OGLE's photometry was
202: used, bringing the chip 2 $V$ photometry into the standard system.
203: In $I$, chip 2 was calibrated differentially with respect to
204: chip 4 using overlapping images obtained on the IAC80 telescope
205: at Iza\~{n}a, Tenerife, Spain, during a photometrical observing run.
206: The dispersion at the center of mass is of the order of 0.01.
207: Hence, the total error in our photometry is that given by \citet{uda01},
208: i.e., up to 0.02 for both $V$ and $I$ bands.
209: Figure~\ref{CMD} shows our final ($V$-$I$, $I$) CMD\footnote[5]{The
210: photometric data are available from the first author upon request.}.
211: The spatial distribution of these stars is presented in
212: Fig.~\ref{star_distrib}.
213:
214:
215: \begin{figure}
216: \epsscale{0.8} % preprint
217: \epsscale{1.1} % emulateapj
218: \plotone{f2.eps}
219: \figcaption{Distribution of resolved stars in IC1613.
220: North is up and East to the left.
221: \label{star_distrib}}
222: \end{figure}
223:
224:
225: The errors given by ALLFRAME are the residuals of the PSF-fitting
226: procedure, so they should be considered internal errors. Signal-to-noise
227: limitations, stellar crowding, blending and starloss, which we can refer to
228: as observational effects, are important error sources and significantly
229: modify the CMD shape and stellar density distribution \citep{apa95}.
230: To estimate the observational effects and the completeness of our
231: photometry, we resorted to artificial stars tests.
232: See \citet{apa95} for a detailed description of the procedure
233: and the effects of crowding.
234: Basically, a large number of artificial stars covering the same range in
235: color and magnitude as the observed stars is added to the images using the
236: corresponding PSF. These were placed on the images following a triangular
237: grid in order to avoid crowding between artificial stars
238: themselves and to optimally sample the chip fields. The photometry is then
239: repeated in the exact same way as was done originally. A comparison of the
240: input and output artificial-star list gives information about the
241: completeness as a function of magnitude and galactocentric distance.
242:
243:
244: \begin{figure}
245: \epsscale{1.0} % preprint
246: \epsscale{1.25} % emulateapj
247: \plotone{f3.eps}
248: \figcaption{{\it Left:} Synthetic CMD produced by IAC-STAR showing the
249: 153260 stars injected in chip 4 ({\it see text for details}).
250: {\it Right:} CMD of the $\sim$40000 {\it artificial} stars recovered both
251: in V and I and considered to have good photometry as described in
252: section 2. The line shows the 25\% completeness limit as determined from the
253: artificial star tests.
254: \label{inj_rec_stars}}
255: \end{figure}
256:
257:
258: As the artificial star sample, we took a synthetic CMD produced by
259: IAC-STAR\footnote[6]{The code, available for free use, can be executed
260: at the IAC-STAR website http://iac-star.iac.es/.} \citep{apa04}
261: using the stellar evolution library of \citet{ber94} and the
262: bolometric corrections of \citet{cas03}. The SFR was chosen
263: constant between 13~Gyr ago and now, while the metallicity range increases
264: from 0.0008 to 0.002 at $t=0$, to 0.0008 to 0.006 at the present time.
265: These ranges were chosen wide enough so that they include the actual
266: metallicity of IC1613. In total, 76630 artificial stars
267: were added in 5 runs in each external chip, and twice as many in
268: the central chip (\#4) where crowding is more important.
269: Figure~\ref{inj_rec_stars} presents the injected and recovered CMDs for
270: chip 4. The broken line indicates the 25\% completeness limit, averaged
271: over the galactocentric radius range, obtained from the ratio of recovered
272: to injected stars as a function of magnitude. Figure~\ref{err} shows the
273: completeness and error in recovered magnitude as a function of input
274: magnitude for an inner and an outer field. It is important noticing that,
275: even though the recovered artificial stars were filtered using the same
276: values of $\sigma$, $\chi ^2$ and $SHARP$ as the real stars, {\it external}
277: errors can be as large as $\sim$1~magnitude at the faint limit.
278:
279:
280: \begin{figure}
281: \epsscale{0.8} % preprint
282: \epsscale{1.1} % emulateapj
283: \plotone{f4.eps}
284: %\plotone{f4_bw.eps}
285: \figcaption{Completeness ({\it left}) and errors in recovered magnitudes
286: ({\it right}) for a central field (chip~4, {\it top panel}), and an outer
287: field (chip~1, {\it bottom panel}). The errorbars in the right panels show
288: the dispersion per magnitude bin.
289: \label{err}}
290: \end{figure}
291:
292:
293: \section{The Color-Magnitude Diagram}\label{descCMD}
294:
295: The most evident feature of our CMD, Fig.~\ref{CMD}, is the red giant
296: branch (RGB), composed of low mass stars
297: \citep[M~$\la 1.8-2.0M_{\Sun}$;][]{chi92} older than about 1~Gyr and
298: burning hydrogen in a thin layer around an inert helium core.
299: The age-metallicity degeneracy and the presence of the fainter extension
300: of the asymptotic giant branch (AGB) in the same region of the CMD makes it
301: difficult to get detailed information about the stars populating it.
302: However, it is possible to place rough limits in time for given
303: metallicities using theoretical isochrones.
304: A more detailed determination of the chemical enrichment law (CEL) using
305: this method, as well as its limitations, is presented in section
306: \ref{recentstarfoMet}. \\
307: The presence of a well populated main sequence (MS) blueward of $V-I=0$ and
308: up to $I \sim 18.5$ indicates very recent star formation ($\la$10~Myr). \\
309: The stars with $0 \leq (V-I) \leq 0.6$ and $I \leq 22$ are most likely blue
310: loop (BL) stars, highlighting the blue edge of the core He-burning loop,
311: while the red edge (i.e., the red supergiant branch, RSG), although
312: contaminated by foreground stars, is well defined from $(V-I) \sim 1.0$ to
313: $(V-I) \sim 1.8$ and up to about $I \sim 16.5$.
314: These are young and intermediate- to high-mass stars, so
315: they are among the most metal-rich stars in IC1613. However, this
316: is a poorly understood phase in stellar evolution and the theoretical
317: models still contain large uncertainties owing to the importance of
318: processes such as mass-loss, overshooting and rotation in very massive
319: stars \citep[see e.g.,][]{mae01}. The BL and RSG actually reach the red
320: clump down to $I=23.76$ \citep{col99}, but the dispersion at the faint
321: end of our CMD makes it impossible to distinguish them from the RGB stars
322: below $I=22$.\\
323: The final extension of the AGB, or red-tail \citep{gal94}, extends
324: horizontally redward from the RGB tip at $I \sim 20$. AGBs are shell H- and
325: He-burning stars, of low and intermediate mass and age over about 0.1~Gyr.
326: Their relatively large number indicates a possibly important
327: intermediate-age population with relatively high metallicity, which would
328: be compatible with the enhanced SFR between 3 to 6~Gyrs ago found by
329: \citetalias{ski03}.
330:
331:
332: \begin{figure}
333: \epsscale{0.48} % preprint
334: \epsscale{1.0} % emulateapj
335: \plotone{f5.eps}
336: \figcaption{Color-magnitude diagram of IC1613 for different galactocentric
337: distances. The radius ranges are
338: {\it (a)}~$r \leq 10\arcmin$,
339: {\it (b)}~$10\arcmin < r \leq 16\farcm5$,
340: {\it (c)}~$r > 16\farcm5$.
341: \label{3CMDs}}
342: \end{figure}
343:
344:
345: Because of the wide field of view, the diagram also contains a relatively
346: large number of foreground stars.
347: Most stars with $0.5 \le V-I \le 1.2$ and $I \le 20$ are
348: probably Galactic dwarfs since their number is the same for the middle
349: and bottom panels in Fig.~\ref{3CMDs} after correcting for the
350: difference in area. More generally, the foreground contamination is
351: rather important for $V-I \ga 0.6$.
352:
353: Although old stars ($\ga$10~Gyr) are very likely present in the RGB,
354: our CMD is not deep enough to detail the old, low metallicity population
355: as it would if fainter MS stars were observed. The existence of a {\it bona
356: fide} old population in IC1613 is known thanks to the presence of RR Lyrae
357: stars \citep{sah92,dol01} and CMDs from Hubble Space Telescope observations
358: showing core helium burning, horizontal branch stars
359: \citepalias{col99,dol01,ski03} and the oldest MS turnoffs \citep{gal07}.
360:
361:
362: \section{Morphology, Spatial Extent \& Distribution of Stellar Populations}\label{morpho}
363:
364: To characterize the morphology of IC1613, we plotted the RGB star
365: distribution obtained from our photometry, and convolved the resulting
366: map with a Gaussian of $\sigma=50 \arcsec$ using IRAF's GAUSS from the
367: IMFILTER package. The result of this process is a smooth map of the
368: stellar density highlighting the morphology of the galaxy. Fitting
369: ellipses to the isodensity contours was done with IRAF's ISOPHOTE routine.
370:
371: From the best fitted ellipses, where the crowding is low but
372: the star number sufficient, we find a position angle of 80$\degr$ and an
373: eccentricity $\epsilon=1-b/a=$ 0.15, in good agreement with the
374: values given by \citet[$PA=81 \degr$, $\epsilon=0.18$;][]{abl72}.
375: Following the shape of the isopleths, we divided the galaxy into concentric
376: ellipses at small radii ($\la$10$\farcm5$) and circles at larger ones
377: that we used for radial star counts and stellar populations gradient
378: analysis. Their semi-major axis increases in steps of 100 pixels,
379: corresponding to $33 \arcsec$.
380:
381:
382: \begin{figure}
383: \epsscale{0.8} % preprint
384: \epsscale{1.1} % emulateapj
385: \plotone{f6.eps}
386: \figcaption{Number density of stars as a function of galactocentric distance
387: before ({\it triangles}) and after ({\it diamonds}) background subtraction.
388: The right-hand vertical axis gives a rough estimate of the surface magnitude,
389: calibrated as described in text.
390: The horizontal line is the weighted mean density of ellipses 29 to 41, which
391: has been adopted as the background level. The exponential fit has a scale
392: length of $2\farcm9\pm 0.1$. The vertical dotted line shows where the
393: foreground stars start to dominate, as obtained from the CMDs (see section
394: \ref{morpho}).
395: \label{nb_density}}
396: \end{figure}
397:
398:
399: Figure~\ref{nb_density} shows the radial profile of the galaxy constructed
400: from the number of stars in each ellipse after correcting for completeness.
401: The correction was obtained from the ratio of the number of recovered
402: to injected stars in the artificial star test for each annular region.
403: The area of the ellipses has been calculated via Monte-Carlo sampling,
404: carefully taking into account the gaps between the chips as well as the
405: regions around saturated stars when calculating the effective surface.
406: An approximate surface brightness scale, shown on the right-hand side of
407: Fig.~\ref{nb_density}, was calculated from the
408: stellar density in each ellipse. The calibration was calculated by
409: comparing the star number with the total, sky subtracted flux
410: for each ellipse then averaged over the radius range for consistency.
411: An exponential least square fit to this curve between $2\arcmin$ and
412: $15\arcmin$ from the center gives a scale length of
413: $2\farcm9 \pm 0.1$ ($620\pm20$ pc) and central surface brightness
414: $\mu_{V,0} = 22.7 \pm 0.6$. The former value is slightly smaller
415: than the $760\pm50$ pc ($\sim$3$\farcm5\pm0.2$) given by \citet{hod91}.
416: However, a similar larger length is obtained when fitting only the inner
417: $\sim$7$\arcmin$, as was done by the authors because of the limited depth
418: of their photometry, and omitting the crowding correction.
419: The profile seems to get steeper at $r \sim 12\arcmin$, but this change
420: of slope could be an artifact of the background subtraction and small
421: number statistics.
422:
423: %\subsection{Age Gradient}\label{agegradient}
424:
425:
426: \begin{figure}
427: \epsscale{0.56} % preprint
428: \epsscale{1.09} % emulateapj
429: \plotone{f7.eps}
430: %\plotone{f7_bw.eps}
431: \figcaption{Spatial distribution of young (MS+BL+RSG; {\it top}) and older
432: (AGB+RGB; {\it bottom}) stars.
433: The red ellipse shows the extent of the MS stars (r=10\arcmin).
434: The red dashed circle indicates where the foreground stars
435: start to dominate (r=16\farcm5).
436: The neutral hydrogen emission contours (0.2, 6.2 and 14.2 Jy km s$^{-1}$)
437: from \citet{hof96} have been overplotted in blue ({\it see text for details}).
438: \label{star_distPop}}
439: \end{figure}
440:
441:
442: The large field of view covered by the WFC permits to study the spatial
443: gradients of the stellar population across the galaxy. Ideally, that would
444: give us hints on its formation and evolution. However, in the case of
445: shallow CMDs, the spatial variations of the morphology of the CMD only
446: reflect accurately differences in the SFH over the last several hundred
447: million years.
448: Such variations have been observed in all the known dwarf irregular
449: galaxies through CMD morphology (e.g., WLM: \citealt{min96}; NGC 6822:
450: \citealt{bat06}; Leo A: \citealt{van04}; Phoenix: \citealt{mar99}) or
451: distribution of the variable star populations (e.g., in Phoenix:
452: \citealt{gal04}; in Leo I: \citealt{bal04}).
453:
454: Figure~\ref{3CMDs} shows that it is also the case for IC1613 and
455: confirms the difference in relative number of young and old stars found
456: by \citetalias{ski03} between their outer HST field and a central field
457: studied earlier by \citet{col99}. We divided the galaxy into three regions
458: following the morphology of the CMD of each individual ellipse defined
459: above: the inner part of the galaxy where stars younger than about
460: 500~Myr are present ($r \leq 10\farcm1$, Fig.~\ref{3CMDs}a), the region
461: at intermediate distance with no young star but still a well defined RGB
462: ($10\farcm1 < r \leq 16\farcm5$, Fig.~\ref{3CMDs}b), and the outermost
463: part of our observed field, dominated by foreground stars
464: ($r > 16\farcm5$, Fig.~\ref{3CMDs}c). The corresponding spatial limits,
465: as well as the H\,\textsc{I} contours from \citet{hof96}, are displayed
466: over the stellar distribution of the young and old populations in
467: Fig.~\ref{star_distPop}.
468: Note that the H\,\textsc{I} contours were shifted by $-0\farcm6$ (+100
469: pixels in y) and $2\farcm9$ ($-$500 pixels in x) in right ascension and
470: declination, respectively. This corresponds to the offset between the
471: astrometry of \citet{lak89} and \citet{hof96}, and adopting the former
472: for the H\,\textsc{I} contours correctly places the H\,\textsc{II} regions
473: of \citet{hod90} on top of the star forming regions, and fits our star
474: distribution better.
475: Although the RGB of Fig.~\ref{3CMDs}c is not clearly defined, a substantial
476: fraction of the displayed stars probably belong to IC1613. This
477: corroborates the results of \citet{alb00}, who found carbon stars
478: extending up to $15\arcmin$ from the center of the galaxy, and shows
479: that IC1613 is actually more extended than previously thought.
480:
481:
482: \begin{figure}
483: \epsscale{0.8} % preprint
484: \epsscale{1.15} % emulateapj
485: \plotone{f8.eps}
486: %\plotone{f8_bw.eps}
487: \figcaption{Stellar densities of the different populations of resolved
488: stars after correction for completeness and background contamination.
489: The stars selected for each population are shown in the inset.
490: The RGB profile was divided by four to fit on the graph.
491: The errorbars have been omitted for clarity.
492: Note the difference in the scale length of the older (RGB, AGB) and younger
493: (MS, BL+RSG) populations ({\it see text for details}).
494: \label{pop_densities}}
495: \end{figure}
496:
497:
498: To give a quantitative measure of the gradient in the age composition of
499: the stars in IC1613, the stellar surface
500: density for different populations of resolved stars is presented in
501: Fig.~\ref{pop_densities}. The age gradient is clearly
502: visible: while the density of the older stars follows the expected
503: exponential decrease from the central region, that of the young population
504: peaks at a radius of $\sim$3$\arcmin$ and vanishes rapidly as the radius
505: increases. This results in the scale length of the young population being
506: much smaller than that of the older stars: a fit to the profiles
507: between $3\farcm5$ and $12\farcm1$ from the center of IC1613 gives scale
508: lengths of $1\farcm19 \pm 0.04$, $2\farcm0 \pm 0.1$, $3\farcm2 \pm 0.2$
509: and $3\farcm8 \pm 0.1$ for the MS, BL+RSG, AGB and RGB, respectively.
510: The off-centered peak in the distribution of the young population, visible
511: at $r\sim3\arcmin$--$3\farcm5$ in Fig.~\ref{pop_densities}, is due
512: to the fact that the star forming regions are distributed in a somewhat
513: circular pattern at this distance from the center \citep{hod90}, where
514: \citet{sil06} observed a higher H \textsc{I} column-density.
515:
516:
517: %\subsection{Metallicity Gradient}\label{metgradient}
518:
519:
520: \begin{figure}
521: \epsscale{0.8} % preprint
522: \epsscale{1.1} % emulateapj
523: \plotone{f9a.eps}\\
524: %\plotone{f9a_bw.eps}\\
525: \plotone{f9b.eps}
526: \figcaption{{\it Top:} Width of the RGB of IC1613 at $M_I=-3.5$ as a function
527: of galactocentric distance, before ({\it thin line}) and after ({\it thick
528: red line}) correction for the effects of crowding. The dashed line indicates
529: the mean value of the corrected width beyond $\sim$$5 \arcmin$.
530: {\it Middle:} Width of the injected ({\it long-dashed line}) and recovered
531: ({\it full thin line}) RGB of the artificial stars tests.
532: The thick red line is a fit to the recovered width.
533: {\it Bottom:} Width of an artificial RGB as a function of the age of the
534: youngest stars in the CMD. The small change between $\tau=0$ and $1$~Gyr is
535: probably due to the presence of stars from the RSG sequence. In all cases the
536: errorbars correspond to the dispersion of twenty solutions.
537: The vertical scale is the same on all three panels.
538: \label{RGBwidth}}
539: \end{figure}
540:
541:
542: Because of the large dependence of the RGB color on metallicity, the
543: position of the RGB in a CMD can be used to estimate the metallicity of
544: the stellar system using empirical relations \citep[e.g.,][]{dac90}. For
545: the same reason, the width of the RGB is often considered to be an
546: indication of the metallicity dispersion.
547: On the CMDs presented in Fig.~\ref{3CMDs}, it appears that the RGB in
548: the central region is wider than that at larger radii and suggests the
549: existence of a metallicity gradient.
550: However, a RGB width gradient may also reflect an observational-effects
551: gradient. To check this possibility, the following procedure was used,
552: including measurements of real- and artificial-star RGB width and error
553: estimates. We measured the width of the RGB at $M_I=-3.5$ as a function of
554: galactocentric distance for the real stars as follows:
555: for each radius interval, chosen so that the RGB contains about 800 stars,
556: a subsample of 450 RGB stars was selected at random.
557: The resulting RGB was sliced in intervals of 0.25 magnitude between
558: $20.5 \leq I \leq 22$, and the width of the RGB in each magnitude range
559: was obtained from the full-width at half-maximum (FWHM) of a Gaussian fit
560: to the color function (CF). The width at $M_I=-3.5$ was then interpolated
561: from a linear fit to the FWHM versus magnitude plot. We repeated this
562: operation twenty times using a different subsample of RGB stars each time,
563: and used the dispersion of these solutions as the errorbars. The same
564: process was followed for the synthetic CMD used in the artificial stars
565: tests, except that the radius bins
566: and random subsamples contained 2500 and 1500 stars, respectively.
567: The resulting plot is presented in Fig.~\ref{RGBwidth} for IC1613
568: ({\it top}) and the artificial stars ({\it middle panel}).
569:
570: It shows that the RGB recovered in the artificial star tests is wider
571: in the center of the galaxy. In addition to the expected FWHM
572: widening due to signal-to-noise limitations, crowding in the central part
573: further disperses the stars on the CMD. Crowding is important and affects
574: the width of the RGB up to $r\sim 10 \arcmin$.
575: To correct the observed RGB for the effects of crowding, we calculated
576: the broadening parameter as a function of radius by subtracting in
577: quadrature the injected width from the recovered one. The corrected width
578: was then obtained by subtracting in quadrature the broadening parameter
579: from the observed RGB width, and is shown as the thick red line in the
580: upper panel of Fig.~\ref{RGBwidth}.
581: The FWHM of the corrected RGB still presents a significant
582: variation across the central $\sim$6$\arcmin$.
583: A higher metallicity dispersion in the central region could be responsible
584: for this residual width excess. However, the lower panel of
585: Fig.~\ref{RGBwidth}, showing the width of artificial RGBs for which the
586: age of the youngest stars is progressively older, indicates that this
587: excess is consistent with the presence of stars between 1 and 3--4 Gyr old
588: in the center.
589: Thus, there is no strong evidence of a metallicity gradient across the
590: galaxy but its presence cannot be ruled out with the present set of data.
591:
592:
593: \section{Recent Star Formation History}\label{recentstarfo}
594:
595:
596: \subsection{The Method}\label{recentstarfoMet}
597:
598: The reconstruction of the recent SFH has been performed in a way similar
599: to that described in Hidalgo \& Aparicio (2007, in preparation) or
600: \citet{gal99}. The main difference here resides in the fact that our CMD
601: is not sufficiently deep to contain discriminating information about
602: intermediate-age and old stars. We limited our study to the recent SFH,
603: i.e., the last $\sim$300~Myr, based on the MS, BL and RSG populations.
604: It thus limited the determination of the SFH to the inner
605: r~$\la 10\arcmin$ since these populations are absent beyond this radius.
606:
607:
608: \begin{figure}
609: \epsscale{0.8} % preprint
610: \epsscale{1.2} % emulateapj
611: \plotone{f10.eps}
612: %\plotone{f10_bw.eps}
613: \figcaption{CEL of IC1613. The diamonds indicate, for a given age, the
614: minimum and maximum values found by fitting theoretical isochrones from the
615: library of Padua \citep{gir02} to the upper RGB. For reference, the CELs
616: obtained via the `Tolstoy' and `Dolphin methods' from \citetalias{ski03} are
617: shown as red circles and blue triangles, respectively.
618: The grayed region is the CEL used for the comparison CMD in the SFH analysis.
619: The stripped region in the last 300~Myr shows the metallicity range that
620: gave the lowest $\chi^2$ when determining the SFH.
621: The right-hand scale was converted from the abundance by mass assuming
622: $Z_\Sun=$~0.0198.
623: \label{CEL}}
624: \end{figure}
625:
626:
627: In short, the SFH is derived from the comparison of the star distribution
628: of the observed CMD with that of a synthetic CMD.
629: The synthetic CMD to be used in the comparison needs to cover a range of
630: age and metallicity at least as large as the one that can be expected in
631: such a dwarf irregular galaxy. It was generated by IAC-STAR \citep{apa04}
632: using the stellar evolution library of \citet{ber94} and the
633: bolometric corrections from \citet{cas03}, with the following
634: parameters: the SFR was chosen constant between 13~Gyr ago and now, the
635: initial mass function (IMF) was that of \citet{kro93}, and the
636: fraction of binaries was set to zero. To fix the input CEL we
637: fitted isochrones of different metallicities to the upper RGB of the
638: outer field where crowding is not dominant.
639: We used the \citet{gir02} isochrones for different ages (1.585,
640: 2.239, 5.012, 10.00 and 12.59~Gyr old) and metallicities between $Z=0.0005$
641: and $Z=0.006$ in steps of 0.0005. The metallicities that are not available
642: in the original library were interpolated using IAC-STAR.
643: A metallicity was considered valid at a given age if the
644: corresponding isochrone was inside the FWHM of the RGB between $20.4 \le
645: I \le 21$. Figure~\ref{CEL} ({\it filled diamonds}) shows the values
646: obtained through this method.
647: Although the method is rather simplistic, our resulting CEL is
648: in very good agreement with those derived by \citetalias{ski03}.
649: However, \citet{gal05} showed that the theoretical isochrones are generally
650: too vertical with respect to the empirical ones, leading to a slight
651: overestimation of the metallicity. The grayed region represents the
652: metallicity range employed to create the comparison CMD.
653: It is mainly used for producing the RGB, from which we obtain the mean SFR
654: between 1500 and 13000~Myr, and serves a normalization purpose for the young
655: SFH.
656: For the stars younger than 300~Myr, we tried to put constraints on their
657: metallicity by further restricting the metallicity range in the comparison
658: CMD to intervals of 0.002 in abundance by mass, and ran the algorithm three
659: times with the following ranges: $Z=$~0.0015-0.0035, $Z=$~0.0025-0.0045
660: and $Z=$~0.0035-0.0055.
661:
662: In order to simulate the observational effects and allow a more realistic
663: comparison with the real CMD, we applied the dispersion in color and
664: magnitude of the synthetic stars recovered from the crowding tests to the
665: comparison CMD, following the same procedure as in \citet{gal99}.
666:
667: In the observed and comparison CMDs, the MS, BL and RSG sequence are
668: then divided into `boxes'. Because the box selection could influence the
669: resulting SFH, we used three different types of box to rule
670: out this possibility: two regular grids with large
671: (0.3 and 0.5 in color and magnitude, respectively) and small (0.2 and 0.3)
672: box size, and an ``\`a la carte'' parametrization. In the latter, the
673: size and shape of the boxes are chosen taking into account
674: the knowledge and model limitations of the evolutionary phases, in
675: particular the slope of the RSG sequence and the position of the BL. The
676: regular grids were also shifted in color and magnitude to check for
677: consistency and ensure the significance of the result.
678:
679: Additionally, the synthetic stars are divided into partial models,
680: each with a small range in age.
681: The temporal resolution is limited by the quantity of information contained
682: in the CMD, which depends on the quality of the data. It was chosen by
683: comparing the capacity of the algorithm to recover the known SFH of a
684: synthetic CMD, after the observational effects had been simulated as
685: described above, using different time ranges.
686: Above a certain time resolution---which depends on age---the recovered
687: SFH was mainly made of short, violent bursts of star formation separated
688: by periods of inactivity, regardless of the input SFH \citep[see
689: also][]{apa07}.
690:
691: The reconstructed SFH is a linear combination of the different partial
692: models. The best solution is obtained through $\chi^2$-minimization by
693: a genetic code. A thorough description of the algorithm and method is
694: presented in \citet{apa07}.
695: In total, we used 24 models with different CMD parametrization and
696: time resolution. The consistency of the method was checked by
697: solving the SFH of synthetic CMDs and comparing the solutions with the
698: input SFHs.
699: Some of the regular grid models did not give a satisfactory solution, but
700: the discrepancies could generally be traced to small differences in the
701: location of evolutionary phases in the CMD, while the solutions obtained
702: with the {\it \`a la carte} parametrization were found to be more stable
703: and are the ones we will present here.
704:
705:
706: \subsection{The Results}
707:
708:
709: \begin{figure*}
710: \epsscale{0.75} % preprint
711: \epsscale{0.80} % emulateapj
712: \plotone{f11.eps}
713: %\plotone{f11_bw.eps}
714: \figcaption{Best SFR(t) obtained for each galactocentric distance.
715: The radial ranges were chosen so that the corresponding CMDs contain the
716: same number of stars. The error bars correspond to the dispersion of twenty
717: solutions for which $\chi^2_\sigma$ = $\chi^2+1$, where $\chi^2$ is the
718: residual of the best solution, indicated in each panel. The gap in the
719: solutions between 300 and $\sim$$1500$~Myr is due to the lack of
720: information from either the young populations or the RGB in this age range.
721: The solutions obtained by \citet{col99} and \citet{ski03} for their
722: respective WFPC2 field are plotted in panels {\it (a)} and {\it (e)}.
723: \label{SFH}}
724: \end{figure*}
725:
726:
727: Of the three metallicity ranges used for the stars younger than 300~Myr,
728: the second one, i.e., $Z=$~0.0025-0.0045, gave the best solutions, and is
729: shown as a stripped region in Fig.~\ref{CEL}. This is consistent with the
730: H\,\textsc{II} region metallicity of \citet{tal80}.
731: Our best solutions for the SFHs at different galactocentric distances are
732: presented in Fig.~\ref{SFH}. The radius ranges were chosen so that the CMDs
733: contain the same number of stars.
734: The error bars correspond to the dispersion of twenty solutions for which
735: $\chi^2$ = $\chi^2_{best}+1$ \citep[see][]{apa07}.
736:
737: The overall picture is a relatively constant SFR at all radii, decreasing
738: with increasing radius, while the mean age of the stars increases with
739: radius.
740: The very central region, panel {\it (a)}, shows a fairly constant SFR
741: for the last 300~Myr of
742: $(1.6 \pm 0.8) \times 10^{-3} $~M$_\Sun~$yr$^{-1}~$kpc$^{-2}$,
743: in excellent agreement with the value found by \citet[$1.6\times 10^{-3}
744: $~M$_\Sun~$yr$^{-1}~$kpc$^{-2}$;][]{col99} using the $V$-band luminosity
745: function of their WFPC2 central field and assuming a Salpeter IMF. The
746: sharp drop in the last $\sim$25~Myr is in agreement with the lack of very
747: bright stars at the center of IC1613 first noted by \citet{hod91}.
748:
749: Between $2\farcm5$ and $\sim$6$\arcmin$ [{\it (b)-(d)}] the SFR
750: increased by a factor $\sim$2-3 in the last 100-150~Myr. This corresponds
751: to the peak in the radial distribution of young stars and H\,\textsc{II}
752: regions found by \citet{hod90}.
753:
754: The field studied by \citetalias{ski03} is a small fraction
755: of the region for which the SFH is represented in panel {\it (e)}.
756: The SFH calculated via the `Cole method'\footnote[7]{Three methods were
757: used in \citetalias{ski03} to calculate the SFH: the `Dolphin method'
758: gives the SFR on a relative scale only.
759: For the `Tolstoy method', the vertical scale of their Fig.~7 gives a mean
760: SFR~$\sim3\times 10^{-3}$~M$_\Sun~$yr$^{-1}~$kpc$^{-2}$, a factor
761: $\sim$15--20 higher than that of the `Cole method' and as high as the mean
762: value of the SFR over the whole galaxy \citep{mat98}.}
763: in \citetalias{ski03}, shown as the
764: green long-dashed lines, is very similar to the one we obtained here for
765: the whole elliptical annulus.
766: At larger radii, the number of stars formed more recently than about
767: 300~Myr decreases to a negligible value when the radius reaches
768: $\sim$10$\arcmin$.
769:
770:
771: \section{Conclusions}\label{concl}
772:
773: We have presented an analysis of the stellar content, morphology,
774: and recent star formation history of the Local Group dIrr galaxy IC1613
775: based on wide-field ($V$-$I$,$I$) photometry of resolved stars.
776:
777: The distribution of resolved stars can be fitted with ellipses of
778: position angle 80$\degr$ and eccentricity 0.15. The exponential fit of the
779: resulting radial profile has a scale length of $2\farcm9 \pm 0.1$.
780: The relatively large number of RGB stars still present in the outer part of
781: our observed field ($r > 16\farcm5$) indicates that the galaxy is actually
782: more extended than previously estimated.
783:
784: The well-populated young evolutionary phases of the CMD of the central
785: region of IC1613 indicate very recent star formation ($\la$10~Myr).
786: The changing CMDs as a function of galactocentric distance show a strong
787: gradient in the age of the younger stellar population, with the young stars
788: lying preferentially in the central part, while the older ones are distributed
789: more uniformly. No evidence of star formation more recent than about
790: 300~Myr was detected beyond $r \ga 10\arcmin$.
791:
792: Analysis of the width of the RGB as a function of radius is consistent
793: with no metallicity gradient. The combination of crowding effect and the
794: presence of younger stars in the RGB is responsible for the widening toward
795: the center of the galaxy.
796:
797: In the region where the recent SFH could be studied ($r \la 10\arcmin$),
798: the results indicate a decreasing SFR(t) from the center outward as
799: expected from the distribution of neutral gas in IC1613, with the
800: exception of the annular region where the star forming regions are clustered
801: ($r\sim4\arcmin$) and therefore the SFR is a factor $\sim$2-3 higher.
802:
803:
804: \acknowledgments
805:
806: We would like to thank Dr.~A. Udalski and the OGLE collaboration for sharing
807: their $VI$ photometry of IC1613, and the anonymous referee for valuable
808: comments.
809: This research project has been supported by a Marie Curie Early Stage Research
810: Training Fellowship of the European Community's Sixth Framework Programme under
811: contract number MEST-CT-2004-504604, the IAC (grant P3/94) and the Spanish
812: Education and Science Ministry (grant AYA2004-06343).
813: As the reviewing of this article was near an end, our friend and coauthor
814: Maurizio passed away in a tragic accident. His ideas and personality will
815: be missed by many. %Rest in peace, my friend.
816:
817:
818: \begin{thebibliography}{}
819:
820: \bibitem[Ables(1972)]{abl72} Ables, H.~D.\ 1972, Publications
821: of the U.S.~Naval Observatory Second Series, 20, 1
822:
823: \bibitem[Albert et al.(2000)]{alb00} Albert, L., Demers, S.,
824: \& Kunkel, W.~E.\ 2000, \aj, 119, 2780
825:
826: \bibitem[Aparicio \& Gallart(1995)]{apa95} Aparicio, A.,
827: \& Gallart, C.\ 1995, \aj, 110, 2105
828:
829: \bibitem[Aparicio \& Gallart(2004)]{apa04} Aparicio, A.,
830: \& Gallart, C.\ 2004, \aj, 128, 1465
831:
832: \bibitem[Aparicio \& Hidalgo(2007)]{apa07} Aparicio, A., \& Hidalgo,
833: S. L. 2007, \aj, submitted
834:
835: \bibitem[Battinelli et al.(2006)]{bat06} Battinelli, P.,
836: Demers, S., \& Kunkel, W.~E.\ 2006, \aap, 451, 99
837:
838: \bibitem[Bertelli et al.(1994)]{ber94} Bertelli, G., Bressan,
839: A., Chiosi, C., Fagotto, F., \& Nasi, E.\ 1994, \aaps, 106, 275
840:
841: \bibitem[Borissova et al.(2004)]{bor04} Borissova, J.,
842: Kurtev, R., Georgiev, L., \& Rosado, M.\ 2004, \aap, 413, 889
843:
844: \bibitem[Baldacci et al.(2004)]{bal04} Baldacci, L.,
845: Matonti, F., Rizzi, L., Clementini, G., Held, E.~V., Momany, Y.,
846: Di Fabrizio, L., \& Saviane, I.\ 2004, Memorie della Societa
847: Astronomica Italiana, 75, 126
848:
849: \bibitem[Castelli \& Kurucz(2003)]{cas03} Castelli, F., \&
850: Kurucz, R.~L.\ 2003, IAU Symposium, 210, 20P
851:
852: \bibitem[Chiosi et al.(1992)]{chi92} Chiosi, C., Bertelli,
853: G., \& Bressan, A.\ 1992, \araa, 30, 235
854:
855: \bibitem[Cole et al.(1999)]{col99} Cole, A.~A., et
856: al.\ 1999, \aj, 118, 1657
857:
858: \bibitem[Da Costa \& Armandroff(1990)]{dac90} Da Costa,
859: G.~S., \& Armandroff, T.~E.\ 1990, \aj, 100, 162
860:
861: \bibitem[Dolphin et al.(2001)]{dol01} Dolphin, A.~E.,
862: et al.\ 2001, \apj, 550, 554
863:
864: \bibitem[Gallart et al.(1994)]{gal94} Gallart, C., Aparicio, A., Chiosi, C.,
865: Bertelli, G., \& Vilchez, J.~M.\ 1994, \apjl, 425, L9
866:
867: \bibitem[Gallart et al.(1999)]{gal99} Gallart, C., Freedman,
868: W.~L., Aparicio, A., Bertelli, G., \& Chiosi, C.\ 1999, \aj, 118, 2245
869:
870: \bibitem[Gallart et al.(2004)]{gal04} Gallart, C., Stetson,
871: P.~B., Hardy, E., Pont, F., \& Zinn, R.\ 2004, \apjl, 614, L109
872:
873: \bibitem[Gallart et al.(2005)]{gal05} Gallart, C., Zoccali,
874: M., \& Aparicio, A.\ 2005, \araa, 43, 387
875:
876: \bibitem[Gallart(2007)]{gal07} Gallart, C. 2007, in IAU Symp. 241,
877: Stellar Populations as Building Blocks of Galaxies, ed. A. Vazdekis \&
878: R. Peletier (Cambridge: Cambridge Univ. Press), in press
879:
880: \bibitem[Girardi et al.(2002)]{gir02} Girardi, L., Bertelli,
881: G., Bressan, A., Chiosi, C., Groenewegen, M.~A.~T., Marigo, P.,
882: Salasnich, B., \& Weiss, A.\ 2002, \aap, 391, 195
883:
884: \bibitem[Hodge et al.(1990)]{hod90} Hodge, P., Lee, M.~G.,
885: \& Gurwell, M.\ 1990, \pasp, 102, 1245
886:
887: \bibitem[Hodge et al.(1991)]{hod91} Hodge, P.~W., Smith,
888: T.~R., Eskridge, P.~B., MacGillivray, H.~T., \& Beard, S.~M.\ 1991,
889: \apj, 369, 372
890:
891: \bibitem[Hoffman et al.(1996)]{hof96} Hoffman, G.~L.,
892: Salpeter, E.~E., Farhat, B., Roos, T., Williams, H., \& Helou, G.\ 1996,
893: \apjs, 105, 269
894:
895: \bibitem[Kroupa et al.(1993)]{kro93} Kroupa, P., Tout,
896: C.~A., \& Gilmore, G.\ 1993, \mnras, 262, 545
897:
898: \bibitem[Lake \& Skillman(1989)]{lak89} Lake, G. R., \& Skillman, E. D. 1989,
899: \aj, 98, 1274
900:
901: \bibitem[Lee et al.(2003)]{lee03} Lee, H., Grebel, E.~K.,
902: \& Hodge, P.~W.\ 2003, \aap, 401, 141
903:
904: \bibitem[Lee et al.(1993)]{lee93} Lee, M.~G., Freedman,
905: W.~L., \& Madore, B.~F.\ 1993, \apj, 417, 553
906:
907: \bibitem[Lozinskaya et al.(2003)]{loz03} Lozinskaya, T.~A.,
908: Moiseev, A.~V., \& Podorvanyuk, N.~Y.\ 2003, Revista Mexicana de
909: Astronomia y Astrofisica Conference Series, 15, 284
910:
911: \bibitem[Maeder \& Meynet(2001)]{mae01} Maeder, A., \&
912: Meynet, G.\ 2001, \aap, 373, 555
913:
914: \bibitem[Magrini et al.(2005)]{mag05} Magrini, L., et al.\
915: 2005, \mnras, 361, 517
916:
917: \bibitem[Mart{\'{\i}}nez-Delgado et al.(1999)]{mar99}
918: Mart{\'{\i}}nez-Delgado, D., Gallart, C., \& Aparicio, A.\ 1999, \aj,
919: 118, 862
920:
921: \bibitem[Mateo(1998)]{mat98} Mateo, M.~L.\ 1998, \araa, 36, 435
922:
923: \bibitem[Minniti \& Zijlstra(1996)]{min96} Minniti, D.,
924: \& Zijlstra, A.~A.\ 1996, \apjl, 467, L13
925:
926: \bibitem[Pietrzy{\'n}ski et al.(2006)]{pie06}
927: Pietrzy{\'n}ski, G., Gieren, W., Soszy{\'n}ski, I., Bresolin, F.,
928: Kudritzki, R.-P., Dall'Ora, M., Storm, J., \& Bono, G.\ 2006, \apj, 642,
929: 216
930:
931: \bibitem[Saha et al.(1992)]{sah92} Saha, A., Freedman, W.~L.,
932: Hoessel, J.~G., \& Mossman, A.~E.\ 1992, \aj, 104, 1072
933:
934: \bibitem[Silich et al.(2006)]{sil06} Silich, S., Lozinskaya,
935: T., Moiseev, A., Podorvanuk, N., Rosado, M., Borissova, J., \&
936: Valdez-Gutierrez, M.\ 2006, \aap, 448, 123
937:
938: \bibitem[Skillman et al.(1989)]{ski89} Skillman, E.~D.,
939: Kennicutt, R.~C., \& Hodge, P.~W.\ 1989, \apj, 347, 875
940:
941: \bibitem[Skillman et al.(2003)]{ski03} Skillman, E.~D.,
942: Tolstoy, E., Cole, A.~A., Dolphin, A.~E., Saha, A., Gallagher, J.~S.,
943: Dohm-Palmer, R.~C., \& Mateo, M.\ 2003, \apj, 596, 253
944:
945: \bibitem[Stetson(1987)]{ste87} Stetson, P.~B.\ 1987, \pasp,
946: 99, 191
947:
948: \bibitem[Stetson(1994)]{ste94} Stetson, P.~B.\ 1994, \pasp,
949: 106, 250
950:
951: \bibitem[Talent(1980)]{tal80} Talent, D.~L.\ 1980,
952: Ph.D.~Thesis
953:
954: \bibitem[Udalski et al.(2001)]{uda01} Udalski, A.,
955: Wyrzykowski, \L., Pietrzy\'nski, G., Szewczyk, O., Szyma\'nski, M.,
956: Kubiak, M., Soszy\'nski, I., \& \.Zebru\'n, K.\ 2001, Acta Astronomica,
957: 51, 221
958:
959: \bibitem[van den Bergh(2000)]{ber00} van den Bergh, S.\ 2000, The
960: galaxies of the Local Group (Cambridge, UK: Cambridge University Press)
961:
962: \bibitem[Vansevi{\v c}ius et al.(2004)]{van04} Vansevi{\v c}ius, V.,
963: et al.\ 2004, \apjl, 611, L93
964:
965: \end{thebibliography}
966:
967:
968: \end{document}
969:
970: