1: \nofiles

2: \documentclass[english,12pt]{article}

3:           \usepackage{babel}

4:           \usepackage[T1]{fontenc}

5:           \usepackage[latin2]{inputenc}

6:           \usepackage{pslatex}

7: \textwidth 17cm \oddsidemargin -.4cm \textheight 24cm

8: \topmargin -1.7cm

9: 

10: \begin{document}

11: 

12: 

13: \title{\bf Multiperiodic RR~Lyrae Stars in

14: Omega Centauri}

15:     \author{P. Moskalik$^1$ ~and~ A. Olech$^1$}

16: \date{$^1$ Copernicus Astronomical Center, Bartycka 18, 00-716 Warszawa, Poland}

17: \maketitle

18: 

19: \begin{abstract}

20: 

21: We have conducted a systematic search for multiperiodic pulsations in

22: the RR~Lyrae-type stars of the galactic globular cluster $\omega$~Cen.

23: Secondary periodicities close to the primary pulsation frequency have

24: been detected in 17 out of 70 studied fundamental mode (RRab) pulsators

25: and in 35 out of 81 overtone (RRc) pulsators. Because of the observed

26: period ratios, these newly detected periodicities must correspond to

27: nonradial modes. Their beating with the primary radial pulsation leads

28: to a slow amplitude and phase modulation, commonly referred to as the

29: Blazhko effect. The incidence rate of Blazhko modulation in

30: ${\omega}$~Cen RRab stars ($24\pm 5\%$) is similar to that observed in

31: the Galactic Bulge. In the case of $\omega$~Cen RRc stars, the incidence

32: rate of Blazhko effect is exceptionally high ($38\pm5\%$), more than 3

33: times higher than in any other studied population.

34: 

35: In addition to Blazhko variables, we have also identified two

36: RR~Lyr variables exhibiting first overtone/second overtone

37: double-mode pulsations, and a triple-mode High Amplitude

38: $\delta$ Scuti variable.

39: 

40: \end{abstract}

41: 

42: 

43: \section{Introduction}

44: 

45: Omega Centauri ($\omega$ Cen) is the largest globular cluster of the

46: Galaxy. It contains about one million stars of which almost 500 are known

47: to be variable (Kaluzny et al. 2004, Weldrake et al. 2007). Recently,

48: precise $B$ and $V$ CCD photometry of variables in $\omega$ Cen were

49: obtained by Cluster AgeS Experiment (CASE) team (Kaluzny et al. 2004).

50: Among others, they published precise light curves of 151 RR Lyr stars

51: belonging to the cluster. These light curves contained from 594 to 761

52: points and were collected in period from 1999 February 6/7 to 2000

53: August 9/10.

54: 

55: \section{Frequency analysis}

56: 

57: We conducted the search for periodicities in individual RR Lyr stars.

58: The analysis was performed in two ways.

59: 

60: We searched for additional signal in $\omega$~Cen RR~Lyrae stars

61: with the standard consecutive prewhitening technique. First, the

62: data were fitted with the single frequency Fourier sum of the form:

63: 

64: \begin{equation}

65: m(t) = A_{0} + \sum_{k} A_{k} \cos (2\pi k f_{0} t + \phi_{k})

66: \end{equation}

67: 

68: The pulsation frequency $f_{0} = 1/P_{0}$ was also optimized in the

69: fitting process. The fit Eq.(1) was then subtracted from the data

70: and the residuals were searched for secondary periodicities. This

71: was done with the Fourier power spectrum computed over the range of

72: $0-5$ c/d. If any additional signal was detected, a new Fourier

73: fit with all frequencies identified so far and with their linear

74: combinations was performed. All frequencies were optimized anew. The

75: new fit was subtracted from original time-series data and residuals

76: were searched for additional periodicities again. The process was

77: repeated until no new frequencies appeared. At this stage, we

78: performed data clipping, by rejecting all measurements deviating

79: from the fit by more than 5$\sigma$, where $\sigma$ is the

80: dispersion of residuals. After removing deviating datapoints, the

81: entire frequency analysis was repeated.

82: 

83: The second method, also based on the consecutive prewhitenings,

84: differed by treatment of the residua. We started from the ANOVA

85: periodogram (Schwarzenberg-Czerny 1996) of the raw light curve and

86: searched it for the peak frequency. Then the light curve was prewihtened

87: with the frequency just identified. Depending on the amplitudes we

88: removed up to 15 harmonics of the frequency from the the {\em

89: prewhitened data} of the previous stage. Next we recomputed the ANOVA

90: periodogram for the residuals and the whole procedure was repeated

91: recursively till no feature exceeded ANOVA value of 15.

92: 

93: It is worth to note that both methods produce, within errors, the same

94: results. The only discrepancies appeared when a real period and its alias

95: have similar height and one method preferred one peak and the second

96: the other one. In this case we have chosen the fit with smaller formal

97: error.

98: 

99: \section{Results}

100: 

101: We detected secondary periodicities close to the primary pulsation

102: frequency in 17 out of 70 fundamental mode pulsators (RRab or RR0

103: stars) and in 31 out of 81 first overtone pulsators (RRc or RR1

104: stars). The new frequencies are well-resolved within our dataset

105: and do not result from secular period variability. The derived

106: period ratios are incompatible with those of the radial modes.

107: Consequently, the newly detected secondary frequencies must

108: correspond to nonradial modes of oscillations. The beating between

109: these nonradial modes and the primary radial pulsation leads to a

110: slow modulation of the observed light curve, a phenomenon known as

111: the Blazhko effect.

112: 

113: \vspace{7.6cm}

114: 

115: \special{psfile=f.ps hoffset=80 voffset=-100 vscale=55 hscale=55

116: angle=0}

117: 

118:    \begin{figure}[h]

119:       \caption{\sf 

120: Typical examples of different

121: frequency patterns encountered in the RRab variables. A true monoperiodic 

122: star is shown in the top panel of the plot.}

123: \end{figure}

124: 

125: In RRab variables two nonradial modes were detected in most cases.

126: Together with the primary peak, they form an equally spaced

127: triplet of frequencies, centered on the primary peak. Following

128: Alcock et~al. (2000) we call such stars RR0-BL variables. In 3

129: RRab stars only one nonradial mode is present (RR0-$\nu1$

130: variables). Finally, an even more complicated pattern was detected

131: in variable V11. In this object we found an equidistant triplet

132: plus an additional peak on the high frequency side of the primary

133: peak. In Fig.\thinspace 1 we display typical examples of different

134: frequency patterns encountered in the RRab variables. For

135: comparison, we show a true monoperiodic star in the top panel of

136: the plot.

137: 

138: 

139: In RRc variables usually only one nonradial mode was detected

140: (RR1-$\nu1$ stars). Equidistant triplets were found only in 4

141: variables (RR1-BL stars). However, more complicated patterns are

142: seen in RRc stars much more frequently than in the RRab stars. In

143: 6 variables we detected {\it nonequidistant} triplets of modes. In

144: other 4 RRc stars, even richer frequency patterns with up to 7

145: nonradial modes were found.

146: 

147: Our findings are summarized in Tables\thinspace 1 and 2. In

148: $\omega$~Cen fundamental mode RR~Lyrae stars the incidence

149: rate of nonradial modes is $24.3\pm 5.1$\%. This is

150: twice the rate observed in the LMC, but identical to the rate in

151: the Galactic Bulge. For $\omega$~Cen overtone RR~Lyrae stars

152: the incidence rate of nonradial modes is $38.3\pm 5.4$\%, which is

153: exceptionally high. This is by far the highest rate among all

154: studied RRc populations, being 3 times higher than in the Bulge

155: and 4 times higher than in the LMC.

156: 

157: 

158: \bigskip

159: 

160: \noindent {\sc Table 1: Incidence rates of nonradial modes in RRab

161: stars in LMC, Galactic Bulge and $\omega$ Cen.}

162: \smallskip

163: \begin{center}

164: \begin{tabular}{|l|c|c|c|c|}

165: \hline

166: \hline

167:  & LMC & Bulge & Bulge & $\omega$ Cen \\

168:  & Alcock et al. & ~~~~Mizerski~~~~ & Collinge et al. & ~~~this work~~~ \\

169:  & (2003) & (2003) & (2006) & \\

170: \hline

171: \hline

172: RR0 & 6158 & 1942 & 1888 & 70 \\

173: \hline

174: RR0-$\nu 1$ & ~400 \hfill 6.5\%~ & ~243 \hfill 12.5\%~ & ~167 \hfill 8.8\%~ & ~~~3 \hfill 4.3\%~ \\

175: RR0-BL & ~331 \hfill 5.4\%~ & ~143 \hfill 7.4\%~ & ~282 \hfill 14.9\%~ & ~13 \hfill 18.6\%~ \\

176: RR0-oth & ~~~20 \hfill 0.3\%~ & ~~~86 \hfill 4.4\%~ & ~~~75 \hfill 4.0\%~ & ~~~1 \hfill 1.4\%~ \\

177: \hline

178: All NR & ~751 \hfill 12.2\%~ & ~472 \hfill 24.3\%~ & ~524 \hfill 27.8\%~ & ~17 \hfill 24.3\%~ \\

179: \hline

180: \hline

181: \end{tabular}

182: \end{center}

183: \bigskip\medskip

184: 

185: 

186: \noindent {\sc Table 2: Incidence rates of nonradial modes in RRc

187: stars in LMC, Galactic Bulge and $\omega$ Cen.}

188: \smallskip

189: \begin{center}

190: \begin{tabular}{|l|c|c|c|c|}

191: \hline

192: \hline

193:  & LMC & LMC & Bulge & $\omega$ Cen \\

194:  & Alcock et al. & Nagy \& Kov\'acs & ~~~Mizerski~~~ & ~~this work~~ \\

195:  & (2000) &  (2006)          & (2003) & \\

196: \hline

197: \hline

198: RR1 & 1143 & 1161 & 771 & 81 \\

199: \hline

200: RR1-$\nu 1$ & ~24 \hfill 2.1\%~ & ~~~46 \hfill 4.0\%~ & ~22 \hfill 2.9\%~ & ~17 \hfill 21.0\%~ \\

201: RR1-BL & ~28 \hfill 2.4\%~ & ~~~53 \hfill 4.6\%~ & ~30 \hfill 3.9\%~ & ~~~4 \hfill 4.9\%~ \\

202: RR1-oth & ~~~8 \hfill 0.7\%~ & ~~~13 \hfill 1.1\%~ & ~41 \hfill 5.3\%~ & ~10 \hfill 12.3\%~ \\

203: \hline

204: All NR & ~60 \hfill 5.2\%~ & ~112 \hfill 9.6\%~ & ~93 \hfill 12.1\%~ & ~31 \hfill 38.3\%~ \\

205: \hline

206: \hline

207: \end{tabular}

208: \end{center}

209: \bigskip

210: 

211: \section{Other interesting cases}

212: 

213: In the course of systematic frequency analysis of $\omega$~Cen

214: RR~Lyrae stars we found two double mode pulsators. These are the

215: first double mode RR~Lyrae variables identified in this globular

216: cluster. The period ratios in both stars are close to 0.80, which

217: implies pulsations in the first two radial overtones. Properties

218: of both variables are summarized in Table\thinspace 3.

219: \bigskip

220: 

221: \noindent {\sc Table 3: FO/SO double mode RR~Lyrae stars in $\omega$~Cen}

222: \smallskip

223: \begin{center}

224: \begin{tabular}{|l|c|c|c|c|c|}

225: \hline

226: \hline

227: Star &  $P_1$ [day] &  $P_2$ [day] &  $P_2/P_1$ & $A1$ [mag] & $A2$ [mag]\\

228: \hline

229: V10   & ~0.3749761~ & ~0.2991768~ & ~0.797856~ & ~0.18305~ & ~0.00668~ \\

230: V145  & ~0.3732140~ & ~0.2986517~ & ~0.800216~ & ~0.20481~ & ~0.00894~ \\

231: \hline

232: \hline

233: \end{tabular}

234: \end{center}

235: \bigskip

236: 

237: RR~Lyrae stars pulsating in the first and second overtone (FO/SO

238: variables) have been previously discovered in the LMC (Alcock

239: et~al. 2000). However, Soszy\'nski et~al. (2003) have noticed,

240: that all these stars are about 1\thinspace mag brighter than

241: typical RR~Lyrae stars in the LMC and suggested that they might

242: belong to short-period double mode Cepheids. In contrast, the two

243: FO/SO double mode pulsators discovered in $\omega$~Cen belong to

244: the RR~Lyrae population of the cluster without any doubt.

245: 

246: In one object, V168, we discovered triple mode pulsations. This

247: star varies with $P_0=0.3212979$\thinspace day,

248: $P_1=0.2465336$\thinspace day and $P_2=0.1967626$\thinspace day.

249: The resulting period ratios are $P_1/P_0=0.767305$ and

250: $P_2/P_1=0.798117$. These ratios indicate that V168 pulsates in

251: the fundamental mode and the first two radial overtones (FU/FO/SO

252: pulsator). The value of $P_1/P_0$ is too large for RR~Lyrae stars,

253: but it is rather typical for High Amplitude $\delta$~Scuti stars

254: (HADS), especially for those with longest periods (Poretti et~al.

255: 2005). We note in passing, that period ratios of V168 are very

256: similar to ratios observed in two other triple mode HADS, V829~Aql

257: (Handler et~al. 1998) and V823~Cas (Jurcsik et~al. 2006).

258: 

259: 

260: 

261: \bigskip

262: \section*{References}

263: 

264: \noindent Alcock C., Allsman R., Alves D.R., et al., 2000, {\sl ApJ}, 

265: {\bf 542}, 257

266: 

267: \noindent Alcock C., Alves D.R., Becker A.,  et al., 2003, {\sl ApJ}, 

268: {\bf 598}, 597

269: 

270: \noindent Collinge M., Sumi, T., Fabrycky D., 2006, {\sl ApJ}, {\bf 651}, 197

271: 

272: \noindent Handler G., Pikall H., Diethelm R., 1998, IBVS 4549

273: 

274: \noindent Jurcsik J., Szeidl B., V\'aradi M., et~al., 2006, {\sl Astron.

275: Astrophys.}, {\bf 445}, 617.

276: 

277: \noindent Kaluzny J., Olech A., Thompson I.B., et al.,  2004, {\sl Astron. 

278: Astrophys.}, {\bf 424}, 1101

279: 

280: \noindent Mizerski T., 2003, {\sl Acta Astron.}, {\bf 53}, 307

281: 

282: \noindent Nagy A., Kov\'acs G., 2006, {\sl Astron. Astrophys.}, {\bf 454}, 257

283: 

284: \noindent Poretti E., Su\'arez J.C., Niarchos P.G., et~al., 2005,

285: {\sl Astron. Astrophys.}, {\bf 440}, 1097.

286: 

287: \noindent Schwarzenberg-Czerny A., 1996, {\sl ApJ Letters}, {\bf 460}, L107

288: 

289: \noindent Soszynski I., Udalski A., Szymanski M. et al., 2003, {\sl Acta Astron.},

290: {\bf 53}, 93

291: 

292: \noindent Weldrake D.T.F., Sackett P.D., Bridges T.J., 2007, {\sl Astron. J},

293: {\bf 133}, 1447

294: 

295: \vspace{0.5cm}

296: \begin{flushleft}

297: {\bf Acknowledgments.} ~This work was supported by MNiI grant

298: number 1 P03D 011 30 to P.M.

299: \end{flushleft}

300: 

301: \end{document}

302: