1: % macho1.tex -- The Distance of the RR Lyrae Variables in the LMC
2:
3: \documentclass[12pt,preprint]{aastex}
4: \def\ltorder{\mathrel{\raise.3ex\hbox{$<$}\mkern-14mu
5: \lower0.6ex\hbox{$\sim$}}}
6: \def\day{{}$^{\rm d}$\llap{.}}
7: \def\msun{{M_\odot}}
8: \def\lsun{{L_\odot}}
9: \def\deg{\mathrel{\raise.9ex\hbox{$\circ$}\mkern-7mu
10: \lower0.4ex\hbox{$\cdot$}}}
11:
12:
13: \begin{document}
14:
15: \title{The Distance of the First Overtone RR Lyrae Variables
16: in the MACHO LMC Database: A New
17: Method to Correct for the Effects of Crowding}
18:
19: \author{C. M. Clement, X. Xu and A. V. Muzzin}
20: \affil{Department of Astronomy and Astrophysics,
21: University of Toronto, Toronto, ON, Canada, M5S 3H8}
22:
23: \begin{abstract}
24:
25: Previous studies have indicated that many of the
26: RR Lyrae variables in the LMC have properties similar to the ones in
27: the Galactic globular cluster M3.
28: Assuming that the M3 RR Lyrae variables
29: follow the same relationships among period, temperature,
30: amplitude and Fourier phase parameter $\phi_{31}$ as their
31: LMC counterparts, we have used the
32: M3 $\phi _{31}-\log P$ relation to identify the
33: M3-like unevolved first overtone RR Lyrae variables in 16
34: MACHO fields near the LMC bar.
35: The temperatures of these variables were calculated from the
36: M3 $\log P - \log T_{eff}$ relation
37: so that the extinction could be derived for each star separately.
38: Since blended stars have lower amplitudes for a given period,
39: the period-amplitude relation should be a useful tool for
40: determining which stars are affected by crowding.
41: We find that the low amplitude LMC RR1 stars are brighter than
42: the stars that fit the M3 period-amplitude relation and
43: we estimate that at least 40\% of the stars are blended.
44: Simulated data for three of the crowded stars illustrate that an
45: unresolved companion with $V \sim 20.5$ could account for the
46: observed amplitude and magnitude.
47: We derive a corrected mean apparent magnitude
48: $\langle V_0 \rangle =19.01 \pm 0.10$ (extinction) $\pm 0.02$
49: (calibration) for the 51 uncrowded
50: unevolved M3-like
51: RR1 variables. Assuming that the unevolved RR1 variables
52: in M3 have a mean absolute magnitude $M_V=0.52 \pm 0.02$
53: leads to an LMC distance modulus $\mu = 18.49 \pm 0.11$
54: \end{abstract}
55:
56: \keywords{galaxies: Magellanic clouds - distances and redshifts -
57: stars: variables: other}
58:
59: \section{Introduction}
60: \label{sec-introduction}
61:
62: Variable stars are useful standard candles for determining
63: the distances to nearby galaxies, but
64: crowding can be a major source of uncertainty in the measurement of
65: these distances.
66: If the image of a variable is blended with that of another
67: star, its apparent magnitude will be too bright. As a result,
68: its derived distance will be too
69: small. It is not always possible to recognize which stars
70: are affected by the crowding
71: so the problem is usually addressed by taking a statistical
72: approach.
73: Another consequence of image blending is that
74: the amplitude of light variation is reduced.
75: Thus, if one can determine which
76: variables have low amplitudes for their periods, the blended
77: stars can be flagged.
78: In this paper, we propose a method, based on the period-amplitude
79: relation, for identifying crowded stars in the LMC.
80: We test our method on the 330
81: RR Lyrae stars in the MACHO database that we classified as bona fide
82: RR1\footnote{We use the system of notation that Alcock et al. (2000b)
83: introduced for RR Lyrae variables: RR0 for fundamental mode, RR1 for
84: first-overtone and RR01 for double-mode (fundamental and first-overtone),
85: instead of the traditional RRab, RRc and RRd.} variables in our previous
86: study (Clement et al. 2002; Alcock et al. 2004, hereafter referred to
87: as A04).
88:
89: Investigations of the LMC RR Lyrae variables
90: by the MACHO collaboration have indicated
91: that many of them have characteristics similar to the ones in the
92: Galactic globular cluster M3. In a preliminary study that included
93: 500 RR Lyrae stars, Alcock et al. (1996) calculated that the mean period of
94: the RR0
95: variables with a $V$ amplitude of $0.8$ mag was 0\day 552,
96: compared with 0\day 480 for M107,
97: 0\day 507 for M4, 0\day 543 for M3 and M72 and 0\day 617
98: for M15. In a subsequent paper, (Alcock et al. 2000a),
99: a more rigorous selection that included only the least
100: crowded RR0 stars\footnote{The
101: percentage flux inside the point-spread function box contributed by
102: neighboring stars was estimated and only the $\sim 20\%$ least
103: crowded stars were included in their sample.} in 16 fields near the bar
104: was made.
105: The ridgeline of the period-amplitude relation they plotted
106: for the RR0 variables
107: was similar to the M3 ridgeline based on the data of Kaluzny et al.
108: (1998).
109: The population of double-mode variables in the LMC is another
110: feature that indicates a similarity to M3.
111: More than 150 of the 181 RR01 stars discovered by
112: Alcock et al. (1997, 2000b) have fundamental mode
113: periods between 0\day 46 and 0\day 50. The only globular
114: clusters known to have RR01 variables with periods
115: in this range are M3 (Corwin et al. 1999, Clementini et al. 2004)
116: and IC 4499 (Clement et al. 1985, Walker \& Nemec 1996).
117:
118:
119: In this study, we use Simon's (Simon \& Lee 1981)
120: Fourier decomposition technique.
121: Previous investigations have pointed to a relationship among
122: his Fourier phase parameter $\phi_{31}$, and the
123: luminosity and metal abundance.
124: Evidence for this can be seen in
125: Figure \ref{fig-phigc} which shows
126: $\phi_{31}-\log P$ plots for the RR1 variables in
127: four well studied globular clusters:
128: M107, M5, M3 and M68.
129: The diagram illustrates
130: that the RR1 variables in a given cluster show
131: a sequence of $\phi_{31}$ increasing with period and that the lower the
132: cluster metallicity, the more the sequence is shifted
133: to longer periods in the plot. Since
134: the luminosities of RR Lyrae variables are known to be
135: correlated with
136: metal abundance, a plot like
137: Figure \ref{fig-phigc} should be useful for identifying
138: a group of RR1 stars with similar luminosities.\footnote{Simon
139: \& Clement (1993) derived equations relating
140: masses, luminosities and temperatures to pulsation period
141: and $\phi_{31}$ based on hydrodynamic
142: pulsation models.
143: Later Catelan (2004) and Cacciari et al. (2005,
144: hereafter referred to as CCC) both demonstrated that the
145: equations should not be applied to individual stars, but CCC
146: also noted that Fourier parameters could be used
147: for estimating the average luminosity of a group of
148: stars, after careful and proper calibration. In this investigation,
149: we do not use the Fourier parameters to determine physical properties
150: for individual stars.}
151: The $\phi_{31}-\log P$ plot that A04 made for
152: the LMC
153: indicated that most of the RR1 variables are similar to
154: the ones in M2, M3 and M5.
155: We therefore assume that the luminosities of the LMC RR1 variables
156: are comparable to the ones in these three clusters.
157: Since M3 is the most variable-rich globular cluster
158: and also because CCC recently made a detailed multicolor
159: and Fourier study of its RR Lyrae variables,
160: we will compare the LMC stars with the ones in M3.
161: Our modus operandi will be to use $\phi_{31}$ to
162: identify the M3-like RR1 variables
163: in the LMC, then use the period-amplitude relation to
164: select the ones that are uncrowded and
165: apply the M3 distance modulus to derive
166: their absolute magnitudes.
167: The $\phi_{31}$ parameter is effective for this analysis because
168: it is not altered by crowding.
169: Simon \& Clement (1993) performed
170: simulations that added constant light
171: to RR Lyrae light curves and found that $\phi_{31}$
172: remained unchanged.
173: By taking this approach, we avoid using the RR Lyrae
174: $M_V-\rm{[Fe/H]}$ relation and the problems that arise
175: because of the uncertainty of its
176: slope.
177:
178: In \S \ref{sec-identM3like}, we discuss our
179: sample selection. Then in
180: \S \ref{sec-crowd}, we use
181: the CCC period-amplitude relation
182: to identify the crowded stars and
183: perform simulations to ascertain the nature of their unresolved
184: companions.
185: In \S \ref{sec-extinction}, we use
186: CCC's M3
187: period-temperature relation to calculate the temperature for each
188: LMC star so that the interstellar extinction
189: and corrected magnitude $V_0$ can be derived
190: and in \S \ref{sec-tilt}, we consider the effects of the
191: LMC geometry on the apparent magnitude of the RR Lyrae variables.
192: Finally in
193: \S \ref{sec-distance}, we derive an LMC distance.
194:
195:
196:
197: \section{The Analysis}
198: \label{sec-analysis}
199:
200: \subsection{Identification of M3-like variables in the LMC}
201: \label{sec-identM3like}
202:
203: In their seminal study of the M3 RR Lyrae variables, CCC published
204: photometric and Fourier parameters for 23 variables that they
205: classified as type RRc (RR1).
206: We use their $V$ data, but limit our study to the unevolved stars.
207: Thus we exclude V70, V85, V129, V170 and V177.
208: These five stars have longer periods for a given amplitude
209: than the others and appear to have evolved off the ZAHB.
210: CCC called them ``long$P$/overluminous" stars.
211: Another three stars, V105, V178 and V203, are excluded
212: because they have short periods ($P<0.29$ days)
213: and small amplitudes. They do not fit into the same
214: period-amplitude sequence as the other RR1 variables.\footnote{Some
215: of these short period variables could
216: be second overtone (RR2) pulsators.
217: CCC plotted the Fourier parameters $\phi_{21}$
218: versus $A_{21}$ for V105, V178 and V203 and concluded that at least
219: V203 is an RR2 variable.}
220: This leaves 15 `unevolved' RR1 variables in our M3 reference sample.
221: In Figure \ref{fig-M3rr1}, we plot $\phi_{31}$,
222: $A_V$ (the $V$ amplitude), and
223: $\langle V \rangle$
224: against $\log P$ for these stars. The diagram illustrates
225: that $\phi_{31}$ and $A_V$ are both correlated
226: with period, but $\langle V \rangle$ is not. The central lines in
227: the $\phi_{31}$ and
228: $A_V-\log P$ plots are least squares fits to the data. The outer
229: lines have the same slope and are envelope lines that encompass all of
230: the data.
231: It is important that none of these 15 stars
232: is affected by crowding. To check this, we verified that there
233: was no correlation between the observed $V$ magnitude and $\Delta A_V$,
234: the displacement from the central line in the $A_V-\log P$ plot of
235: Figure \ref{fig-M3rr1}. Furthermore, the M3 finding chart published
236: by Bailey (1913) indicates that all of these variables are located
237: outside the cluster core. Therefore it seems reasonable to assume that
238: crowding does not affect our reference sample.
239:
240: Our LMC data are the MACHO data for 330 RR Lyrae
241: variables in 16
242: fields\footnote{A chart showing the locations of
243: the MACHO fields is available at http://www.macho.mcmaster.ca}
244: near the LMC bar. The observations were obtained between 1992
245: and 1999. The 330 stars that we analyse were all classified as
246: bona fide RR1 variables by A04 who published their
247: photometric and Fourier parameters.
248: Figure \ref{fig-LMCperphi} shows the $\phi_{31} - \log P$ plot for these
249: stars, with the envelope lines for M3 superimposed.
250: LMC variables with $\log P > -0.45$ (equivalent to $P=$ 0\day 355)
251: have been excluded from the plot because
252: M3 RR1 variables with periods greater than this appear
253: to have evolved.
254: In order to decide which of the LMC stars to include in our
255: sample, we used a weighting scheme that Gladders \& Yee
256: (2000) devised for determining whether data points belong to a linear
257: sequence, given their error.
258: We constructed a Gaussian distribution for
259: $\phi_{31}$ of each star, using a HWHM equal to the error listed by
260: A04. Then we determined a `weight' by measuring
261: the fraction of the area under the
262: Gaussian that fell between
263: the envelope lines of Figure \ref{fig-LMCperphi}.
264: Figure \ref{fig-wthist} shows
265: the distribution of these weights. For
266: our analysis, we will include only the stars
267: with the highest probablility of fitting the M3 $\phi_{31}-\log P$
268: plot, i.e. the ones with weight greater than $0.5$.
269:
270:
271: \subsection{The Crowded Stars}
272: \label{sec-crowd}
273:
274:
275: \subsubsection{Correction for crowding}
276: \label{sec-corcrowd}
277:
278: The period-$V$ amplitude relation for the M3-like RR1
279: variables in the LMC is shown in Figure \ref{fig-LMCperamp} with
280: the envelope lines from Figure \ref{fig-M3rr1} superimposed.
281: Of the 147 stars plotted, 71 lie below the lines, 54 lie
282: between them and 22 are above. If our hypothesis that low
283: amplitudes are caused by blending is correct, we would expect
284: the stars below the lines to be brighter. However, before
285: we proceed to test this hypothesis, we need to consider
286: the stars with the shortest periods ($\log P< -0.54$, i.e. $P<$ 0\day 29).
287: They all have low
288: amplitudes and could therefore
289: be the LMC counterparts of the M3 stars, V105, V178 and V203.
290: Since their low amplitudes might be intrinsic, they
291: are not suitable for our crowding
292: test. We therefore exclude all LMC variables with
293: $P<$ 0\day 29 from our analysis.
294: This leaves 127 stars: 54 below the lines, 51 between and 22 above.
295: The mean magnitudes for the stars in these three
296: regimes are listed in Table \ref{tbl-magpa1}.
297: Each row represents a different threshold for the weights
298: of the stars considered.
299: We will base our discussion on all of the stars with weight $> 0.5$.
300:
301: The data of Table \ref{tbl-magpa1} demonstrate that the
302: stars that lie below the M3 period-amplitude relation are brighter than the
303: ones that lie between the lines.
304: This is the result we expect if the low amplitude stars are blended.
305: A t-test (for all the stars with weight $>0.5$)
306: indicates that the difference in $\langle V \rangle$ is highly
307: significant, with a probability of only $0.0056$
308: that the two groups of stars are drawn from the same population.
309: As for the stars that lie above the M3 lines, the difference
310: between their mean
311: $\langle V \rangle$ and that of the stars between
312: the lines is not significant.
313: Two M3 counterparts to these stars, V85 and V177,
314: are displaced with respect to the other RR1 variables
315: in the period-amplitude relation,
316: but their $\phi_{31}$ and $\langle V \rangle$ values do not
317: set them apart.
318: CCC classified them with their long period/high luminosity group.
319: We therefore suggest that these high amplitude stars are evolved,
320: even though they do not appear brighter than the others.
321:
322: Thus we can account for
323: the relative mean magnitudes of the stars below, between and above the
324: lines in Figure \ref{fig-LMCperamp}.
325: In this discussion, we have not taken the effect of interstellar
326: extinction into account and we know that it
327: may vary from star to star.
328: However, if the average extinction among the stars in each
329: of the three regimes of Figure \ref{fig-LMCperamp} is the same,
330: the ranking of their mean $\langle V \rangle$ values should be correct.
331:
332: Table \ref{tbl-magpa1} indicates that the $V$ magnitudes
333: of 54 of 127 M3-like RR1 variables appear to be altered by crowding.
334: This represents approximately 40\% of the stars.
335:
336:
337: \subsubsection{Crowding Simulations}
338: \label{sec-crowdsim}
339:
340: If the RR1 stars that lie below the envelope lines
341: in the period-amplitude diagram have unresolved
342: companions, what are the apparent magnitudes of these companions?
343: To answer this question, we performed simulations to ascertain how
344: the presence of an unresolved companion would affect the
345: observed
346: magnitude and amplitude. A few examples of these simulations
347: are presented in Table \ref{tbl-crowdsim}.
348: The simulations show the change in $V$ magnitude and amplitude
349: when an RR1 variable with $V=19.4$ or $19.7$ mag
350: is blended with a star with $V=20-22$ mag.
351: This range of $V$ magnitudes was selected
352: for the companions because
353: the LMC color magnitude diagram
354: plotted by Alcock et al. (2000a) shows
355: a high density of main sequence stars with $V>20$ mag and therefore
356: the RR Lyrae are probably blended with stars like these.
357: In Table \ref{tbl-blendsim}, we show simulated data for
358: three stars that
359: are displaced from the central line in Figure \ref{fig-LMCperamp}
360: by more than $0.1$ mag.
361: According to the table, the `true' magnitudes of these
362: RR Lyrae stars could be $V= 19.70$, $19.56$ and $19.63$
363: while their
364: unresolved companions have $V$ magnitudes, $21.25$, $20.28$ and
365: $20.50$ respectively.
366: These RR Lyrae $V$ magnitudes are typical LMC RR1 magnitudes
367: and the $V$ magnitudes of the companions are consistent with
368: LMC main sequence stars that belong to a younger population.
369: RR Lyrae variables belong to an old population for which the main
370: sequence turn-off is about $3.4$ mag fainter than the horizontal
371: branch. Thus older main sequence stars would be too faint to
372: have much of an effect.
373:
374: The MACHO CM diagram also shows a high concentration of RR Lyrae
375: variables and horizontal branch
376: red clump stars with $V\sim 19.2$ mag, but we do not
377: expect the stars in our data set to be blended with stars this
378: bright. LMC RR Lyrae
379: variables that have unresolved companions with $V\sim 19.2$ mag would
380: be brighter than 19th magnitude. Such stars are known to exist in the
381: MACHO database, but the 147 stars in our sample are fainter.
382: In an independent study of LMC RR Lyrae variables,
383: Di Fabrizio et al. (2005, hereafter referred to as
384: DF05)\footnote{A major study of RR Lyrae variables near the LMC bar
385: was made by
386: Clementini et al. (2003, hereafter referred to as C03) and DF05.
387: They made photometric observations
388: with the $1.54$ m Danish telescope at La Silla, Chile and
389: spectroscopic observations with the $3.6$ m ESO telescope and the VLT.
390: They discovered approximately 135 RR Lyrae
391: variables in two fields that overlap with parts of MACHO fields 6 and
392: 13.}
393: identified five stars that could be blended with red clump stars
394: among the RR Lyrae stars in their sample.
395: These five stars had
396: typical RR Lyrae periods, but small amplitudes and bright
397: mean $V$ ($<19$) magnitudes.
398: They concluded that one of these anomalous stars was an RR0 blended
399: with a young main sequence star, but did not assign a definite
400: classification to the other four.
401:
402: \subsection{The Extinction}
403: \label{sec-extinction}
404: A serious difficulty in deriving the distance to LMC stars is
405: that the amount of interstellar extinction is not constant.
406: Schwering \& Israel (1991) estimated that the foreground extinction due
407: to dust in the Galaxy ranges from $E(B-V)=0.07$ to $0.17$ over the
408: LMC surface. Furthermore Harris et al. (1997) concluded that the
409: distribution of dust within the LMC itself is clumpy. Therefore it is
410: desirable to derive the extinction for the stars individually
411: and we are in a position to do this. Since
412: we have selected M3-like stars for our investigation,
413: we can calculate the temperature of each star individually
414: and then derive its reddening.
415: We use CCC's M3 period-temperature
416: relation:
417: \begin{eqnarray}
418: A= 13.353 -1.19 \log P_0 - 4.058 \log T_{eff}.
419: \end{eqnarray}
420: CCC found that a value of $A=-1.82\pm 0.03$
421: gave the best fit for the unevolved variables, but it
422: predicted temperatures that were too low for stars that had
423: evolved away from the ZAHB.
424: Assuming that $A=-1.82$
425: and that
426: $\log P_0 = \log P_1 + 0.127$,\footnote{A typical ratio
427: $P_1/P_0$ for the M3 RR01 variables (Clementini et al. 2004)
428: is $0.746$. This is equivalent to $\Delta \log P=0.127$.}
429: we derive the following equation
430: for calculating the temperatures of the unevolved LMC stars:
431: \begin{eqnarray}
432: \log T_{eff}=3.702-0.293 \log P_1.
433: \end{eqnarray}
434: The
435: unreddened color $(V-R)_0$ can be computed from a relation
436: derived by Kov\'acs \& Walker (1999) based on the models of
437: Castelli, Gratton \& Kurucz (1997a):
438: \begin{eqnarray}
439: \log T_{eff}=3.8997-0.4892(V-R)_0+0.0113 \log g + 0.013 \rm{[M/H]}
440: \end{eqnarray}
441: For this calculation, we assume that [M/H] = $-1.3$, the
442: value adopted by CCC for their M3 study,\footnote{We also note that
443: there is a small, but non-zero metallicity dependence subsumed into CCC's
444: estimation of $A$.}
445: and $\log g =2.93$, the mean of the $\log g$ values they calculated
446: for the 15 stars in our reference sample.
447: In order to calculate the
448: corrected magnitudes $V_0$, we assume
449: a ratio of total to selective absorption:
450: \begin{eqnarray}
451: A_V/E(V-R)=5.35
452: \end{eqnarray}
453: (Schlegel, Finkbeiner \& Davis 1998).
454:
455: The $\langle V \rangle _F$ and $\langle R \rangle _F$
456: magnitudes,\footnote{Our $\langle V \rangle$ and
457: $\langle R \rangle$ values were
458: derived by fitting a 6-order Fourier series of the form:
459: \begin{eqnarray}
460: mag = A_0+\sum _{j=1,6} A_j \cos (j\omega t + \phi _j)
461: \hskip 2mm
462: \end{eqnarray}
463: to the $V$ and $R$ magnitudes for each star,
464: where $\omega$ is ($2\pi$/period). Thus our mean magnitudes are
465: the $A_0$ values from the fit of equation (5) to the observational data.}
466: the $V$ extinction and
467: corrected magnitudes $V_0$ for the 51 uncrowded, unevolved
468: stars are listed in Table \ref{tbl-paramm3}.
469: Figure \ref{fig-LMCexthist} shows the distribution of $E(V-R)$ for
470: the stars that we consider to be uncrowded, unevolved M3-like RR1
471: variables. The mean is $0.073 \pm 0.02$\footnote {A04 estimated
472: that the error in $E(V-R)$ due to uncertainties in the temperature-color
473: transformation would be $\sim 0.01$ mag. In addition, there is an error
474: of $\sim 0.015$ which arises from the uncertainty in the value of
475: $A$ in equation (1).
476: Combining these in quadrature leads to an error
477: of $0.018$ mag in $E(V-R)$ and $0.10$ in the $V$ extinction.}
478: which corresponds to $V$ band extinction of $0.39 \pm 0.10$.
479: This value can be compared with the LMC extinction that
480: Zaritsky et al. (2004) derived
481: using a different technique. They measured effective temperatures
482: and line-of-sight extinction for millions of individual stars
483: by comparing stellar atmosphere models with $U, B, V, I$ photometry.
484: Then they constructed extinction maps for
485: stars in two temperature ranges where the model
486: fitting between temperature and extinction was least degenerate:
487: $5500 K \le T_E < 6500 K$ (cool, older stars) and
488: $12,000 K \le T_E < 45,000 K$
489: (hot, younger stars). They derived a mean $V$ absorption of
490: $0.43$ mag for the cool stars and $0.55$ mag for the hotter stars.
491: With temperatures in the range $\sim 6100$ to $7300 K$, RR Lyrae
492: variables are similar to their cool star group and the mean extinction
493: we have derived agrees with theirs to within our quoted error.
494: However, for the cool stars, Zaritsky et al. found a bimodal distribution
495: which they attributed to the existence of a dust layer.
496: This bimodal structure is not evident in our data, but our sample
497: is several orders of magnitude smaller than theirs.
498: A Shapiro-Wilk W test indicates that our data, plotted
499: in Figure \ref{fig-LMCexthist}, do not deviate significantly from a
500: normal distribution.
501:
502: We can also compare our extinction with the values derived by
503: C03 for their RR Lyrae investigation
504: since the two LMC fields they observed overlap with
505: MACHO fields 6 and 13. They derived $E(B-V) = 0.116\pm 0.017$ for
506: their field A and $0.086\pm 0.017$ for field B from the colors of
507: the edges of the instability strip. The corresponding total
508: $V$ extinction values are $0.36\pm 0.05$ and $0.27\pm 0.05$. The
509: mean extinction we calculate for the 9 stars in MACHO field 6
510: is $0.35$ mag and for the 4 stars in field 13, it is $0.38$.
511: These are in good agreement with the extinction C03 derived
512: for their field A and a bit high compared to their field
513: B value, but still within the quoted errors.
514:
515: We have pointed out that equation (2) predicts temperatures that
516: are too low for stars that have evolved away from the ZAHB.
517: Thus if the stars above the lines in Figure \ref{fig-LMCperamp}
518: have evolved, we would expect the extinction derived from
519: equations (2), (3) and (4) to be underestimated and this would
520: lead to faint $V_0$ values. This is exactly what we see in
521: Table \ref{tbl-magpa2} where we list the mean $V_0$
522: these equations predict.
523: Even though their mean $\langle V \rangle$ is comparable to the mean
524: for the stars between the lines,
525: their $V_0$ is $0.08$ mag fainter.
526: We conclude that this supports our hypothesis that these stars have
527: evolved.
528: We will not include them in the sample of
529: unevolved M3-like RR1 variables we use to determine the LMC
530: distance.
531: For the stars below the lines of Figure \ref{fig-LMCperamp},
532: equation (2) should be valid for computing temperatures, but
533: if they are blended with other stars, their observed $(V-R)$
534: colors are not their true ones. Thus the $V_0$ values
535: we derive for these stars individually will also be in error.
536: However, if the
537: mean extinction for these stars is comparable to the
538: mean extinction for the
539: stars between the lines, they will still appear brighter and this is what
540: Table \ref{tbl-magpa2} illustrates.
541:
542:
543:
544: \subsection{Tilt Correction and Line-of-Sight Distribution}
545: \label{sec-tilt}
546:
547: The distribution of the RR Lyrae population in the LMC has not been
548: well established. Kinematic studies by Freeman et al. (1983),
549: Storm et al. (1991) and Schommer et al. (1992) all indicated
550: that the oldest globular clusters belong to a flattened
551: disk-like system with $\sigma _{RV} \sim 28 \rm{kms}^{-1}$.
552: There was no evidence for the presence of a halo.
553: It was therefore assumed that the RR Lyrae variables must belong to
554: a disk population.
555: However, van den Bergh (2004) pointed out that the
556: observed radial velocities did not rule out the possibility that the globular
557: clusters formed in a halo. Subsequently, radial velocities
558: of LMC RR Lyrae variables derived by Minniti et al. (2003)
559: and by Borissova et al. (2006) have indicated a larger velocity distribution
560: ($\sigma _{RV}= 50 \pm 2 \rm{km s}^{-1}$) and these authors
561: argue that there is an old and metal poor halo in the LMC.
562: If the RR Lyrae variables belong to a halo population, they should
563: be distributed spherically with respect to the LMC center.
564: On the other hand, if they belong to a disk population,
565: the tilt of its plane with respect to the plane of the
566: sky is an effect that must be considered when deriving the distance.
567:
568: We made tilt corrections
569: based on recent investigations of the LMC geometry by van der Marel
570: \& Cioni (2001) and by Nikolaev et al. (2004).
571: Van der Marel \&
572: Cioni (2001) analysed the variations in brightness
573: of asymptotic
574: and red giant branch stars in near-IR color magnitude
575: diagrams extracted from the DENIS and 2MASS surveys.
576: They found a sinusoidal variation in apparent magnitude
577: as a function of position angle, which they
578: interpreted to be the result of distance variations
579: because one side of the LMC plane is closer to us than the
580: other.
581: For their analysis, they
582: assumed that the LMC center is located at $\alpha _0= 82\deg 25$
583: and $\delta _0 = -69\deg 5$
584: (van der Marel 2001) and included
585: stars with $\rho$ in the range $2\deg 5 - 6\deg 7$
586: where $\rho$ is the angular distance from the LMC center.
587: They
588: derived an inclination angle $i = 34\deg 7 \pm 6\deg 2$
589: and line-of-nodes position angle $\Theta = 122\deg 5 \pm 8 \deg 3$.
590: Nikolaev et al. (2004) carried out a similar analysis based
591: on more than 2000 MACHO Cepheids with $\rho < 4^{\circ}$.
592: Assuming $\alpha _0= 79 \deg 4$ and $\delta _0= -69\deg 03$,
593: they derived $i = 30\deg 7 \pm 1\deg 1$
594: and $\Theta = 151\deg 0 \pm 2 \deg 4$.
595:
596: We used the equations listed in \S 2 of van der Marel \& Cioni's paper
597: to calculate corrected $V_0$ magnitudes for both inclinations
598: and they are listed in columns (7) and (8) of
599: Table \ref{tbl-paramm3}.
600: Figure \ref{fig-LMChist} shows the distribution of
601: $V_0$ based on the three different assumptions
602: for the LMC tilt.
603: All three show a peak at approximately $19.0$ mag.
604: Some of the dispersion in $V_0$ is due to depth within the LMC, but
605: we expect a dispersion in absolute magnitude as well because
606: our M3 reference stars have a range of $0.2$ in apparent
607: magnitude.\footnote{The error in $V_0$ due to the error in extinction
608: is $\pm 0.10$ mag. However, this is a systematic error and should not
609: affect the shape of the $V_0$ distributions plotted in Figure
610: \ref{fig-LMChist}.}
611: The normal parameter estimates are listed in Table \ref{tbl-normal}.
612: A Shapiro-Wilk W test indicates
613: that none of the three deviates significantly from
614: a normal distribution.
615:
616: We adopt the tilt corrections
617: based on the viewing angles derived by Nikolaev et al. because
618: the stars in their sample, like ours, are all within
619: $4^{\circ}$ of the LMC center. The mean $V_0$ based on these
620: viewing angles is exactly the same as the value obtained when
621: no tilt correction is applied. Therefore our derived $V_0$ does not
622: depend on any assumption about the distribution of the LMC RR Lyrae
623: variables.
624:
625:
626:
627: \section{The LMC Distance}
628: \label{sec-distance}
629:
630:
631:
632: \subsection{The Apparent Magnitude of the RR1 Variables}
633: \label{sec-appmag}
634:
635: We have adopted a mean $V_0 = 19.02$ for the 51
636: uncrowded, unevolved M3-like RR1 variables, based on the LMC
637: viewing angles derived by Nikolaev et al. (2004).
638: Since our mean magnitudes are the $A_0$ values from
639: equation (5), we need to convert them to intensity means
640: before we compute an LMC distance.
641: A04 showed that the intensity means are $\sim 0.01$ mag brighter
642: than $A_0$ so
643: we revise our adopted mean $V_0$ to $19.01$ mag.
644:
645:
646: To determine the precision of the $V_0$ values listed in Table
647: \ref{tbl-paramm3}, we need to consider
648: the errors in reddening and the errors
649: in the calibration of our photometry.
650: We have already pointed out in \S \ref{sec-extinction} that the error
651: in $E(V-R)$ is $0.018$ mag which corresponds to $0.10$ mag in
652: the $V$ exinction.
653: Our calibrated $V$ and $R$ magnitudes are from the study by A04, and
654: were calculated from transformation
655: equations derived by Alcock et al. (1999),
656: designated calibration version 990318.
657: They derived an internal precision of $\sigma _V = 0.021$ for
658: their $V$ magnitudes
659: by comparing the results for stars in overlapping (MACHO) fields.
660:
661: Another way to test our calibration is to
662: compare with the results from other investigations.
663: Two of the 51 uncrowded, unevolved M3-like RR1 stars that we have
664: listed in Table \ref{tbl-paramm3} were observed by DF05.
665: The mean (intensity) magnitudes derived from the two studies for these
666: stars are listed in Table \ref{tbl-photcomp}
667: and they agree to within $0.01$ mag.
668: In our earlier paper (A04), we listed mean magnitudes for
669: five additional RR1 stars that were observed by both groups and the
670: MACHO magnitudes were significantly brighter.
671: However, in our new analysis, we have classified two of these stars
672: as crowded.
673: Their mean magnitudes are also listed
674: in Table \ref{tbl-photcomp} and it is clear that the MACHO
675: magnitudes are brighter than the ones derived by DF05.
676: The remaining three
677: are not M3-like so we can not ascertain whether or not
678: they are crowded.
679:
680: DF05 also compared their $V$ magnitudes with MACHO data provided
681: by Alves and by Kov\'acs (Alcock et al 2003)
682: and found that the MACHO magnitudes
683: were approximately $0.04$ mag brighter on average.
684: DF05 concluded that the difference
685: occurred because the MACHO reduction procedure did not adequately compensate
686: for crowding in stars with $V> 18.25$. The MACHO collaboration
687: recognized the crowding
688: problem so A04 made crowding corrections by adding artifical
689: stars of known
690: magnitude to the the image frames and measuring $\Delta V$ the
691: difference between their input and recovered magnitudes.
692: However, these corrections
693: introduce a relatively large error of $\sim 0.10$ mag into the adopted
694: apparent magnitudes.
695: In the present investigation, we have dealt with this problem
696: by identifying
697: crowded stars and removing them from our data set.
698:
699: We can not determine which of the MACHO stars in the Alves and
700: Kov\'acs samples
701: were crowded; however, our sample of M3-like stars can provide some insight.
702: According to the data listed in Table \ref{tbl-magpa1},
703: approximately 40\% of the
704: M3-like variables are crowded and as a consequence,
705: their average $V$ magnitude is $0.09$ mag brighter.
706: If we assume that the mean MACHO magnitude for
707: 40\% of the stars in the data sets that Alves and Kov\'acs provided to
708: DF05 is
709: $0.09$ mag brighter (as a result of blending with main sequence stars),
710: while the remaining 60\% have $V$ magnitudes
711: similar to the ones that DF05 derived,
712: we would expect the MACHO stars to be about $0.04$ mag brighter
713: on average. Thus if the crowded
714: stars could be removed from the MACHO sample, the photometry
715: derived in these two independent studies would be in good agreement.
716:
717: Since our $V$ photometry for uncrowded stars appears to be
718: in good agreement with
719: that of DF05, we assume that the uncertainty in our calibrated
720: $V$ magnitudes is $0.02$ mag.
721: Therefore we
722: adopt $V_0 = 19.01 \pm 0.10$ (extinction) $\pm 0.02$ (calibration)
723: for our sample of
724: uncrowded, unevolved M3-like variables in the LMC.
725: This is in good agreement with the result of C03:
726: $V_0= 19.064 \pm 0.064$
727: at [Fe/H]$=-1.5$ on the metallicity scale of Harris (1996).
728: Taking into account the fact that
729: the RR1 variables in our sample are M3-like, that [Fe/H] for M3
730: is $-1.57$ on the Harris scale and that C03 derived $\Delta M_V/\Delta
731: \rm{[Fe/H]}= 0.214$ for the RR Lyrae variables in the
732: LMC, we estimate that $V_0 = 19.05$ at [Fe/H]$=-1.57$
733: for the data of C03.
734:
735: \subsection{The Absolute Magnitude and Distance of the RR1 Variables}
736: \label{sec-absmag}
737:
738: A major challenge in determining the LMC distance
739: is to identify a homogeneous group of stars for which
740: the absolute magnitude is well established.
741: We deal with this problem by using the M3
742: distance modulus to calculate the absolute $V$
743: magnitude of our 15 M3 reference stars. Then we assume that
744: the uncrowded stars in our LMC sample have the same mean $M_V$. The M3 distance
745: modulus has been derived in investigations
746: of RR Lyrae variables by Marconi et al. (2003)
747: and by Sollima et al. (2006).
748:
749:
750: Taking a theoretical approach, Marconi et al.
751: applied pulsation theory
752: to the $BV$ observations of M3 by
753: Corwin \& Carney (2001) and the $K$
754: observations of Longmore et al. (1990).
755: They compared the results from pulsation theory
756: with the observed edges of the instability
757: strip, the observed $K$ band period-magnitude relation
758: and the observed relations among period-magnitude-color and
759: period-magnitude-amplitude.
760: For their calculations,
761: they used bolometric corrections and temperature-color transformations
762: provided by Castelli, Gratton \& Kurucz (1997a, 1997b)
763: and adopted a mean RR Lyrae mass of
764: $0.67 \msun$, based on evolutionary models of Cassisi et al. (2004).
765: They computed a mean distance
766: modulus $DM = 15.07 \pm 0.05$,\footnote{The evolutionary distance modulus that
767: Marconi et al. derived was about $0.08\pm 0.05$ longer than the
768: pulsational value. However, they
769: stated that it would be shorter if element diffusion were properly
770: taken into account because
771: the luminosity of HB models would be about $0.03-0.04$
772: mag fainter.}
773: but pointed out that if they had used the
774: models of Vandenberg et al. (2000) instead,
775: $DM$ would have been $15.05$.
776: We adopt the distance modulus based on the Cassisi models,
777: DM=$15.07 \pm 0.05$.
778: Our 15 M3 reference stars have $V$ magnitudes ranging from
779: $15.48$ to $15.68$ with mean
780: $\langle V \rangle =15.60$. If we apply CCC's extinction,
781: $E(B-V)=0.01 \pm 0.01$, this corresponds to $V_0 = 15.57$
782: which leads to $M_V=0.50 \pm 0.06$.
783:
784: Sollima et al. derived an M3 distance modulus
785: from the RR Lyrae period-metallicity-$K$ band
786: luminosity $(PL_K)$ relation that they calibrated from observations:
787: \begin{eqnarray}
788: M_K=-2.38 (\pm 0.04) \log P_F + 0.08 (\pm 0.11) \rm{[Fe/H]}
789: -1.05 (\pm 0.13)
790: \end{eqnarray}
791: where $M_K$ is the absolute $K$ magnitude, $P_F$ is the fundamentalized
792: pulsation period and [Fe/H] refers to the metallicity scale of Carretta
793: \& Gratton (1997).
794: They derived their
795: coefficient of $\log P_F$ from the slope of the $K-\log P_F$
796: relation for 538 RR Lyrae variables in 16 globular clusters
797: and their [Fe/H] coefficient from the slope of the
798: $(M_K-2.38 \log P)$ - [Fe/H] relation for
799: the four globular clusters in their sample that had
800: distance determinations based on \it Hipparcos \rm trig parallaxes for
801: local subdwarfs. They obtained their zero point from the $K$ magnitude
802: of RR Lyrae combined with the
803: trig parallax that Benedict et al. (2002) measured for it from
804: HST astrometry. From equation (6), Sollima et al. calculated an
805: M3 distance modulus of
806: $15.07$. Thus $M_V=0.50 \pm 0.20$.
807:
808: Unfortunately, the large error in the coefficient
809: of [Fe/H] in equation (6) results in a large uncertainty in $M_K$.
810: Therefore it seems appropriate to calculate the
811: M3 distance modulus directly from the absolute magnitude that
812: Benedict et al. derived for RR Lyrae: $M_V=0.61^{-0.11}_{+0.10}$.
813: The [Fe/H] for RR Lyrae is $-1.39$ (Clementini et al. 1995),
814: which is comparable to the M3 metal abundance, $-1.34$ on the scale
815: of Carretta \& Gratton. Furthermore,
816: the pulsation period of RR Lyrae, 0\day 567, and its maximum $V$
817: amplitude\footnote{RR Lyrae
818: exhibits the
819: Blazhko effect, but according to Szeidl (1988), the maximum
820: light amplitude for a Blazhko star fits the period amplitude relation
821: for singly periodic variables.}
822: which is $0.9$ mag according to
823: Smith et al. (2003) places it on the M3 period-amplitude relation.
824: CCC studied five RR0 stars with periods within $0.01$ days of
825: 0\day 567 (V10, V69, V135, V137 and V142).
826: The mean $V$ amplitude
827: for these stars is $0.89$ mag and their mean $\langle V \rangle$
828: is $15.65 \pm 0.02$. Assuming that their mean absolute magnitude is the same as
829: that of RR Lyrae,
830: we derive a mean $M_V = 0.56\pm 0.11$ for our 15 M3 reference stars.
831:
832: The Baade-Wesselink technique has not been applied to any RR Lyrae
833: variables in M3. However, in a review of RR Lyrae luminosities,
834: Cacciari (2003) reported the result of a B-W analysis for the
835: star RR Cet which has a metal abundance comparable to that of M3.
836: Cacciari et al. (2000) derived $M_V = 0.56 \pm 0.15$ for this star
837: which, with a
838: period of 0\day 553 and $V$ amplitude $0.98$ mag (Simon \& Teays 1982)
839: fits the M3 P-A relation. Six stars analysed by CCC
840: (V36, V40, V71, V89, V133
841: and V149) have periods within $0.01$ days of 0\day 553 and the
842: mean $V$ amplitude and mean $\langle V \rangle$ they derived for these
843: stars were $0.99$ and $15.65$ respectively. By comparing the mean
844: magnitudes of these six
845: stars with those of our 15 M3 reference stars, we derive mean
846: $M_V= 0.52 \pm 0.15$ for the reference stars.
847:
848:
849: The mean of our four $M_V$ values is $0.52$ with a standard deviation
850: $0.02$ mag.
851: Combining this with the
852: $V_0=19.01 \pm 0.10$ (extinction error) $\pm 0.02$ (calibration error)
853: that we derived for the uncrowded, unevolved
854: M3-like RR1 variables in our LMC sample and adding the estimated
855: errors in quadrature, leads to distance
856: modulus $\mu _ {LMC}= 18.49 \pm 0.11$.
857:
858: Our distance modulus is in good agreement with
859: $\mu _0= 18.48 \pm 0.08$ derived by Borissova et al.
860: (2004) from $K$-band photometry of 37 RR Lyrae variables
861: in the inner regions of the LMC. In their investigation,
862: they derived a mean $\langle K
863: \rangle = 18.20$ and assumed that the mean $K$ band absorption
864: $A_K= 0.05$ mag.
865: By following the procedure described by Bono et al. (2001, 2003),
866: they calculated $M_K=-0.332$ at $\log P = -0.30$.
867: Their adopted absolute $K$ magnitude was $0.85$ mag
868: brighter than our adopted $M_V$. This is consistent with
869: the apparent magnitudes we have derived; their mean
870: $K_0$ ($18.15$) is $0.86$ mag brighter than our $V_0$, $19.01$.
871: Our adopted $\mu _0$ is also comparable to the
872: value $(18.48)$ obtained by McNamara et al. (2007) in their recent
873: analysis of $\delta$ Scuti stars.
874:
875: By identifying the crowded
876: stars and removing them from the sample, we have avoided
877: using the crowding corrections that introduced an additional uncertainty
878: of $0.10$ mag to the distance modulus we derived in our
879: earlier study (A04). The main source of error in this investigation
880: is the error in estimating the $V$ extinction.
881:
882:
883: \section{Summary}
884: \label{sec-summary}
885:
886: We have devised a new method for identifying crowded
887: RR1 variable stars in the LMC, based on simulations that show that
888: stars with unresolved companions have low amplitudes
889: for their periods.
890: Given that many LMC RR Lyrae variables have properties similar to the
891: ones in the Galactic globular cluster M3, we used the
892: M3 $\phi_{31}-\log P$ relation to identify the M3-like unevolved
893: RR1 variables in our LMC sample.
894: The Fourier phase parameter $\phi _{31}$ is useful for selecting
895: a homogeneous sample because it is not affected by crowding.
896:
897: When the M3-like variables were plotted on the M3 period-amplitude
898: diagram, we found that the mean $V$ magnitude of the LMC stars
899: with low amplitudes was $0.09$ mag brighter than the mean for the
900: stars that fit the M3 period-amplitude relation.
901: Four of the stars in our sample were observed
902: in the study of LMC RR Lyrae variables by DF05.
903: Comparing the photometry from the two studies, we found that our
904: $V$ magnitudes for the two stars considered to be uncrowded
905: agreed to within $0.01$ mag, while the MACHO $V$
906: magnitudes for the two stars we considered to be crowded were
907: more than $0.05$ mag brighter. From this, we conclude that
908: our method for identifying crowded RR Lyrae variables is effective.
909: It could prove to be useful for identifying crowded RR Lyrae
910: variables in other local group galaxies.
911:
912: We used the M3 period-temperature relation for
913: unevolved RR Lyrae variables to determine the temperature
914: and reddening for each of the uncrowded RR1 variables in our
915: sample. After making corrections for the tilt of
916: the LMC, we derived a mean $V_0$ magnitude of $19.01 \pm 0.10$
917: (extinction) $\pm 0.02$ (calibration).
918: Then to estimate the absolute magnitude, we
919: used the M3 distance modulus and the trig parallax of
920: RR Lyrae to derive the mean absolute magnitude
921: of the unevolved RR1 variables in our M3 reference sample. This
922: turned out to be $M_V= 0.52 \pm 0.02$.
923:
924: Finally, we
925: derived an LMC distance modulus $\mu _{LMC} = 18.49 \pm 0.11$
926: which is in good agreement with the results of other recent studies
927: and
928: with $18.5$ mag, the value employed by the Hubble Space Telescope's
929: key project for measuring the Hubble constant (Freedman et al.
930: 2001).
931:
932:
933:
934:
935:
936: \acknowledgements
937:
938: We thank David Clement and
939: Doug Welch for their helpful comments during the preparation of this
940: manuscript. We also express our gratitude to our referee,
941: Bruce Carney, who made several suggestions that have improved
942: the paper.
943: In addition, financial support from
944: Science and Engineering Research Canada (NSERC) is gratefully
945: acknowledged.
946:
947: \clearpage
948:
949: \begin{thebibliography}{}
950: \bibitem[Alcock et al. 1996]{bib-alc1996} Alcock, C. et al. 1996, \aj,
951: 111, 1146
952: \bibitem[Alcock et al. 1997]{bib-alc1997} Alcock, C. et al. 1997, \apj,
953: 482, 89
954: \bibitem[Alcock et al. 1999]{bib-alc1999} Alcock, C. et al. 1999, \pasp,
955: 111, 1539
956: \bibitem[Alcock et al. 2000a]{bib-alc2000a} Alcock, C. et al. 2000a, \aj,
957: 119, 2194
958: \bibitem[Alcock et al. 2000b]{bib-alc2000b} Alcock, C. et al. 2000b, \apj,
959: 542, 257
960: \bibitem[Alcock et al. 2003]{bib-alc2003} Alcock, C. et al. 2003, \apj,
961: 598, 597
962: \bibitem[Alcock et al. 2004]{bib-alc2004} Alcock, C. et al. 2004, \aj,
963: 127, 334 (A04)
964: \bibitem[Bailey 1913]{bib-bai1913} Bailey, S. I. 1913, Harvard Ann. 78,
965: 1
966: \bibitem[Benedict et al. 2002]{bib-ben2002} Benedict, G. F. et al. 2002, \aj,
967: 123, 473
968: \bibitem[Bono et al. 2001]{bib-bon2001} Bono, G., Caputo, F.,
969: Castellani, V. Marconi, M. \& Storm, J. 2001, \mnras, 326, 1183
970: \bibitem[Bono et al. 2003]{bib-bon2003} Bono, G., Caputo, F.,
971: Castellani, V. Marconi, M., Storm, J. \& Degl'Innocenti, S.
972: 2003, \mnras, 344, 1097
973: \bibitem[Borissova et al. 2004]{bib-bor2004} Borissova, J., Minniti, D.,
974: Rejkuba, M., Alves, D., Cook, K. H., \& Freeman, K. C. 2004 \aap, 423,
975: 97
976: \bibitem[Borissova et al. 2006]{bib-bor2006} Borissova, J., Minniti, D.,
977: Rejkuba, M., Alves, D. 2006 \aap, 460, 459
978: \bibitem[Cacciari 2003]{bib-cac2003} Cacciari, C. 2003, in New
979: Horizons in Globular Cluster Astronomy, ed. G. Piotto, G. Meylan, S. G.
980: Djorgovsli \& M. Riello (ASP Conf. Ser. 296) (San Francisco: ASP), 329
981: \bibitem[Cacciari et al. 2000]{bib-cac2000} Cacciari, C., Clementini,
982: G., Castelli, F. \& Melandri, F. 2000, in IAU Colloq. 176, The Impact
983: of Large-Scale Surveys on Pulsating Star Research, ed. L. Szabados \&
984: D. W. Kurtz (ASP Conf. Ser. 203) (San Francisco: ASP), 176
985: \bibitem[Cacciari et al. 2005]{bib-cac2005} Cacciari, C., Corwin, T. M.
986: \& Carney, B. W. 2005, \aj, 129, 267 (CCC)
987: \bibitem[Carretta \& Gratton 1997]{bib-car1997} Carretta, E. \& Gratton,
988: R. G. 1997, \aaps, 121, 95
989: %\bibitem[Carretta et al. 1998]{bib-car1998} Carretta, E., Cacciari, C.,
990: %Ferraro, F. R., Fusi Pecci, F. \& Tessicini, G. 1998, MNRAS, 298, 1005
991: \bibitem[Cassisi et al. 2004]{bib-cas2004} Cassisi, S., Castellani, M.,
992: Caputo, F. \& Castellani, V. 2004, \aap, 426, 641
993: \bibitem[Castelli, Gratton \& Kurucz 1997a]{bib-cas1997a} Castelli, F.,
994: Gratton, R. G. \& Kurucz, R. L. 1997a, \aap, 318, 841
995: \bibitem[Castelli, Gratton \& Kurucz 1997b]{bib-cas1997b} Castelli, F.,
996: Gratton, R. G. \& Kurucz, R. L. 1997b, \aap, 324, 432
997: \bibitem[Catelan 2004]{bib-cat2004} Catelan, M. 2004, in IAU Colloq. 193,
998: Variable Stars in the Local Group, ed. D. W. Kurtz \& K. R. Pollard
999: (ASP Conf. Ser. 310)
1000: (San Francisco: ASP), 113
1001: \bibitem[Clement et al. 2002]{bib-cle2002} Clement, C. M., Muzzin, A.
1002: V., Rowe, J. F. \& the MACHO Collaboration. 2002, \baas, 34, 651
1003: \bibitem[Clement et al. 1985]{bib-cle1985} Clement, C. M., Nemec, J. M.,
1004: Robert, N., Wells, T., Dickens, R. J. \& Bingham, E. A. 1985,
1005: \aj, 92, 825
1006: \bibitem[Clement \& Shelton]{bib-cle1997}
1007: Clement, C. M. \& Shelton, I. 1997, \aj, 113, 1711
1008: %\bibitem[Clement, Xu \& Muzzin 2005]{bib-cle2005} Clement, C. M., Xu, X. \& Muzzin, A. V. 2005, \baas, 37, 1364
1009: \bibitem[Clementini et al. 1995]{bib-cle1995} Clementini, G., Carretta,
1010: E., Gratton, R. G., Merighi, R., Mould, J. R. \& McCarthy, J.K. 1995,
1011: \aj, 110, 2319
1012: \bibitem[Clementini et al. 2004]{bib-cle2004} Clementini, G., Corwin, T.
1013: M., Carney, B. W. \& Sumerel, A. N. 2004, \aj, 127, 938
1014: \bibitem[Clementini et al. 2003]{bib-cle2003} Clementini, G., Gratton,
1015: R. G., Bragaglia, A., Carretta, E., Di Fabrizio, L. \& Maio, M. 2003,
1016: \aj, 125, 1309 (C03)
1017: \bibitem[Corwin \& Carney 2001]{bib-cor2001} Corwin, T. M. \& Carney,
1018: B.W. 2001, \aj, 122, 3183
1019: \bibitem[Corwin et al. 1999]{bib-cor1999} Corwin, T. M., Carney,
1020: B.W. \& Allen, D. M. 1999, \aj, 117, 1332
1021: %\bibitem[Diaconis \& Efron 1983]{bib-dia1983} Diaconis, P. \& Efron, B.
1022: %1983, Sci. Am., 248, No. 5, 116
1023: \bibitem[Di Fabrizio et al. 2005]{bib-dif2005} Di Fabrizio, L.,
1024: Clementini, G., Maio, M., Bragaglia, A., Carretta, E., Gratton, R.,
1025: Montegriffo, P. \& Zoccali, M. 2005, \aap, 430, 603 (DF05)
1026: \bibitem[Freedman et al. 2001]{bib-fre2001} Freedman, W. L. et al.
1027: 2001, \apj, 553, 47
1028: \bibitem[Freeman et al. 1983]{bib-fre1983} Freeman, K. C., Illingworth,
1029: G., \& Oemler, A. Jr. 1983, \apj, 272, 488
1030: \bibitem[Gladders \& Yee 2000]{bib-gla2000} Gladders, M. D. \&
1031: Yee, H. K. C. 2000, \aj, 120, 2148
1032: \bibitem[Harris et al. 1997]{bib-har1997} Harris, J., Zaritsky, D.,
1033: \& Thompson, I. 1997, \aj, 114, 1933
1034: \bibitem[Harris 1996]{bib-har1996} Harris, W. E. 1996, \aj, 112, 1487
1035: \bibitem[Kaluzny et al. 1998]{bib-kal1998}
1036: Ka{\l}u\.{z}ny, J., Hilditch, R. W., Clement, C., \&
1037: Ruci\'{n}ski, S. M. 1998, \mnras, 296, 347
1038: \bibitem[Kaluzny et al. 2000]{bib-kal2000}
1039: Ka{\l}u\.{z}ny, J., Olech, A., Thompson, I., Pych, W., Krzemi\'{n}ski, W., \&
1040: Schwarzenberg-Czerny, A. 2000, \aaps, 143, 215
1041: \bibitem[Kov\'acs \& Walker 1999]{bib-kov1999} Kov\'acs, G. \&
1042: Walker, A. R. 1999, \apj, 512, 271
1043: \bibitem[Kraft \& Ivans 2003]{bib-kra2003} Kraft, R. P. \&
1044: Ivans, I.I. 2003, \pasp, 115, 143
1045: \bibitem[Longmore et al. 1990]{bib-lon1990} Longmore, A. J., Dixon, R.,
1046: Skillen, I., Jameson, R. F., \& Fernley,
1047: J. A. 1990, \mnras, 247, 684
1048: \bibitem[McNamara et al. 2007]{bib-mcn2007} McNamara, D. H., Clementini,
1049: G. \& Marconi, M. 2007, \aj, 133, 2752
1050: \bibitem[Marconi et al. 2003]{bib-mar2003} Marconi, M., Caputo,
1051: F., Di Criscienzo, M. \& Castellani, M. 2003, \apj, 596, 299
1052: %\bibitem[Marconi \& Clementini 2005]{bib-mar2005} Marconi, M. \&
1053: %Clementini, G. 2005, \aj, 129, 2257
1054: \bibitem[Minniti et al. 2003]{bib-min2003} Minniti, D., Borissova, J.,
1055: Rejkuba, M., Alves, D. R., Cook, K. H., \& Freeman, K.C. 2003, Science,
1056: 301, 1508
1057: \bibitem[Nikolaev et al. 2004]{bib-nik2004} Nikolaev, S., Drake, A. J.,
1058: Keller, S. C., Cook, K. H., Dalal, N., Griest, K., Welch, D. L., \&
1059: Kanbur, S. M. 2004, \apj, 601, 260
1060: \bibitem[Schlegel et al. 1998]{bib-sch1998} Schlegel, D. J., Finkbeiner,
1061: D. P. \& Davis, M. 1998, \apj, 500, 525
1062: \bibitem[Schommer et al. 1992]{bib-sch1992} Schommer, R. A., Olszewski,
1063: E. W., Suntzeff, N. B. \& Harris, H. C. 1992, \aj, 103, 447
1064: \bibitem[Schwering \& Israel 1991]{bib-sch1991} Schwering, P. B. W. \&
1065: Israel, F. P. 1991, \aap, 246, 231
1066: \bibitem[Simon \& Clement 1993]{bib-sim1993} Simon, N. R. \& Clement, C.
1067: M. 1993, \apj, 410, 526
1068: \bibitem[Simon \& Lee 1981]{bib-sim1981} Simon, N. R. \& Lee, A. S.
1069: 1981, \apj, 248, 291
1070: \bibitem[Simon \& Teays 1982]{bib-sim1982} Simon, N. R. \& Teays, T. J.
1071: 1982, \apj, 261, 586
1072: \bibitem[Sollima et al. 2006]{bib-sol2006} Sollima, A., Cacciari, C. \&
1073: Valenti, E. 2006, \mnras, 372, 1675
1074: \bibitem[Smith et al. 2003]{bib-smi2003} Smith, H. A. et al. 2003,
1075: \pasp, 115, 43
1076: \bibitem[Storm et al. 1991]{bib-sto1991} Storm, J., Carney, B. W.,
1077: Freedman, W. L., \& Madore, B. F. 1991, \pasp, 103, 261
1078: \bibitem[Szeidl 1988]{bib-sze1988} Szeidl, B. 1988, in Multimode Stellar
1079: Pulsations, ed. G. Kov\'acs, L. Szabados, \& B. Szeidl (Budapest:
1080: Konkoly Obs.), 45
1081: \bibitem[van den Bergh 2004]{bib-vdb2004} van den Bergh, S. 2004, \aj,
1082: 127, 1897
1083: \bibitem[Vandenberg et al. 2000]{bib-vdb2000} Vandenberg, D. A.,
1084: Swenson, F. J., Iglesias, C. A., \& Alexander, D. R. 2000, \apj,
1085: 532, 430
1086: \bibitem[van der Marel 2001]{bib-vdm2001} van der Marel, R. P.
1087: 2001, \aj, 122, 1827
1088: \bibitem[van der Marel \& Cioni 2001]{bib-vdm2001a} van der Marel, R. P. \&
1089: Cioni, M. L. 2001, \aj, 122, 1807
1090: \bibitem[Walker 1994]{bib-wal1994} Walker, A. R. 1994, \aj, 108, 555
1091: \bibitem[Walker, A. R. \& Nemec 1996]{bib-wal1996} Walker, A. R. \&
1092: Nemec, J. M. 1996, \aj, 112, 2026
1093: \bibitem[Zaritsky et al. 2004]{bib-zar2004} Zaritsky, D., Harris, J.,
1094: Thompson, I. B. \& Grebel, E. K. 2004, \aj, 128, 1606
1095: \end{thebibliography}
1096:
1097: \clearpage
1098:
1099: \begin{figure}
1100: \includegraphics{f1.eps}
1101: \caption{ Plot of $\phi_{31}$ vs. $\log P$ for the RR1
1102: variables in four well studied Galactic globular clusters:
1103: the Oosterhoff type I clusters
1104: M107 (open circles), M5 (solid circles) and M3 (open triangles)
1105: and the Oosterhoff type II cluster M68 (solid triangles).
1106: The $\phi_{31}$ values plotted here
1107: for these four clusters
1108: were determined by Clement \& Shelton (1997), Kaluzny et al.
1109: (2000), Cacciari et al (2005) and
1110: Walker (1994) respectively.
1111: Their metal abundances are $-1.10$, $-1.32$,
1112: $-1.50$ and $-2.43$ on the $\rm{Fe}_{\rm{II}}$ metallicity
1113: scale of Kraft \& Ivans (2003).
1114: \label{fig-phigc}}
1115: \end{figure}
1116:
1117:
1118: \begin{figure}
1119: \includegraphics{f2.eps}
1120: \caption{ Plots of $\phi_{31}$, $V$ amplitude and the (intensity) mean
1121: $V$ magnitude versus $\log P$ for the 15 M3 unevolved RR1 variables
1122: that we include in our sample. The data
1123: are taken from the investigation by CCC.
1124: In the two
1125: upper panels, the central lines represent least squares fits to
1126: the data and the outer lines, plotted with the same slope, are
1127: the envelope lines that encompass all of the data.
1128: \label{fig-M3rr1}}
1129: \end{figure}
1130:
1131: \clearpage
1132:
1133:
1134: \begin{figure}
1135: \includegraphics{f3.eps}
1136: \caption{A plot of the Fourier phase difference $\phi_{31}$ vs
1137: $\log P$ for 330 bona fide RR1 variables in 16 MACHO fields in
1138: the LMC. The superimposed
1139: lines represent the $\phi_{31}- \log P$ relation for the M3
1140: RR1 variables studied by CCC.
1141: (See the upper panel of Figure \ref{fig-M3rr1}.)
1142: \label{fig-LMCperphi}}
1143: \end{figure}
1144:
1145: \clearpage
1146:
1147: \begin{figure}
1148: \includegraphics{f4.eps}
1149: \caption{The distribution of `weights' for the LMC
1150: RR1 variables in our sample.
1151: The weight is a measure of the probability that
1152: a star lies between the envelope lines of Figure
1153: \ref{fig-LMCperphi}.
1154: Our procedure for determining
1155: these weights is described in
1156: \S \ref{sec-identM3like}.
1157: \label{fig-wthist}}
1158: \end{figure}
1159:
1160: \clearpage
1161:
1162:
1163:
1164: \begin{figure}
1165: \includegraphics{f5.eps}
1166: \caption{The period-$V$ amplitude relation for the M3-like RR1
1167: variables in 16 MACHO fields in the LMC. Only
1168: the 147 stars with weight greater than $0.5$ in Figure \ref{fig-wthist}
1169: are included. The superimposed lines represent the
1170: period-amplitude relation for the M3 RR1 variables (see the centre
1171: panel of Figure \ref{fig-M3rr1}).
1172: \label{fig-LMCperamp}}
1173: \end{figure}
1174:
1175: \clearpage
1176:
1177:
1178: \begin{figure}
1179: \includegraphics{f6.eps}
1180: \caption{A histogram of E(V-R) for the 51 uncrowded, unevolved M3-like
1181: RR1 variables in 16 MACHO fields in the LMC.
1182: These are the stars that lie between the envelope lines of
1183: Figure \ref{fig-LMCperamp}.
1184: \label{fig-LMCexthist}}
1185: \end{figure}
1186:
1187:
1188: \clearpage
1189: \begin{figure}
1190: \includegraphics{f7.eps}
1191: \caption{A histogram for the LMC variables
1192: that lie between the lines in Figure
1193: \ref{fig-LMCperamp}. These are the stars that
1194: we consider to be uncrowded M3-like unevolved RR1 variables.
1195: Three histograms
1196: are plotted: the first represents the data with
1197: no tilt correction
1198: and the other two include the tilt corrections of van der Marel
1199: \& Cioni (2001)
1200: and Nikolaev et al. (2004) respectively.
1201: The curves are Gaussian fits to the data.
1202: \label{fig-LMChist}}
1203: \end{figure}
1204:
1205: \clearpage
1206:
1207: \input{tab1.tex}
1208: \input{tab2.tex}
1209: \input{tab3.tex}
1210: \input{tab4.tex}
1211: \input{tab5.tex}
1212: \input{tab6.tex}
1213: \input{tab7.tex}
1214:
1215: \end{document}
1216:
1217: