A&A 449, 569-572 (2006)
DOI: 10.1051/0004-6361:20054191

On the light curve shape of Cepheids in IC 1613 and NGC 6822
(Research Note)

E. Antonello

INAF-Osservatorio Astronomico di Brera, via E. Bianchi 46, 23807 Merate, Italy

Received 12 September 2005 / Accepted 12 December 2005

Abstract
Comparison of the Fourier parameters of fundamental mode Cepheids with period near 10 d in galaxies of the Local Group (IC 1613, NGC 6822, Milky Way, Magellanic Clouds) confirms the previous indication of the lack of a spread of $\phi_{21}$ values in some dwarf irregular galaxies. It is not yet clear whether this is a real effect or if it is just due to the low number of Cepheids in these galaxies. We suspect however that in this period range the Cepheids of IC 1613 and NGC 6822 behave differently from those in the Milky Way and the Magellanic Clouds. The main effect of the different metallicity on the Fourier parameters is confirmed to be the larger R₂₁ values of shorter period Cepheids in metal-poorer galaxies. However the metallicity alone should not be enough to explain the various differences among the Cepheids of the four galaxies. The difference between the spread of $\phi_{21}$ values near 10 d in Milky Way and Magellanic clouds is pointed out.

Key words: stars: variables: Cepheids - galaxies: Local Group - galaxies: stellar content

1 Introduction

In a previous work (Antonello et al. 2000) we tried to verify the possible effects of a very different metallicity on the shape of Cepheids' light curves by comparing stars in IC 1613 with those in the Milky Way. The comparison was made using the Wh-band light curves of Cepheids in IC 1613 (e.g. Mantegazza et al. 2001, and references therein) and V-band light curves of Milky Way stars. The purpose was to offer new observational results in order to verify the predictions of pulsational models, in particular for different metallicities. Resonances among pulsation modes - such as that between the fundamental and the second overtone mode, P₀/P₂=2 for a period $P_0 \sim 10$ d - give rise to observable effects on the light curves, which can be exploited to put constraints on the models. New V- and I-band data are now available for the metal-poor galaxies IC 1613 (Udalski et al. 2001) and NGC 6822 (Pietrzynski et al. 2004), making a better comparison with both Milky Way and Magellanic Cloud Cepheids possible. IC 1613 and NGC 6822 are two dwarf irregular galaxies of the Local Group, with an estimated metallicity of [Fe/H] between -1.3 and -0.7 (Skillman et al. 2003) for IC 1613, and -0.49 (Venn et al. 2001) for NGC 6822. The values of oxygen abundance, $12+\log({\rm O/H})$ , are 7.86 (IC 1613) and 8.14 (NGC 6822), to be compared with 8.02, 8.37, and 8.7 for SMC, LMC, and Milky Way, respectively (van den Bergh 2000).

Table 1: Fourier parameters of the I-band light curves of Cepheids.

Table 2: Fourier parameters of the V-band light curves of Cepheids.

$\begin{figure} \par\includegraphics[width=12.2cm,clip]{4191fig1.ps} \end{figure}$

Figure 1: Comparison of the Fourier parameters of Cepheids in IC 1613 ( triangles) and NGC 6822 ( plus) with those of Cepheids in the Magellanic Clouds (I band photometry; dots) or Milky Way (V band photometry; dots). The dotted line in panels e) and f) was drawn by eye (see text). The error bars are the mean formal error of the Fourier parameters of Cepheids in IC 1613 and NGC 6822, that is, $\pm {0.061}, \pm {0.17}, \pm {0.051}$ , and $\pm {0.13}$ , for $R_{21}(I), \phi _{21}(I), R_{21}(V)$ , and $\phi _{21}(V)$ , respectively.

Open with DEXTER

2 Data analysis

The V and I data of IC 1613 were taken from Udalski et al. (2001); the known period P of pulsation for such stars was slightly improved using other data, as described in Antonello et al. (2006). The V and I data and the P values of NGC 6822 stars were taken from Pietrzynski et al. (2004). Since the Cepheids with the best light curves are those with longer P, we took the stars with $P \ge 5$ d into account. The light curves were Fourier-analyzed (cosine series, e.g. Antonello et al. 2000), and the resulting main parameters are reported in Table 1 (I-band light curve) and Table 2 (V-band light curves). They include the names of Cepheids (those of NGC 6822 with the label "cep'' and those of IC 1613 with the label "ogle''), the logarithm of P, the amplitude ratio R₂₁ and the respective formal uncertainty, the phase difference $\phi_{21}$ , and the formal uncertainty. The very uncertain Fourier parameters (errors in R₂₁ and $\phi_{21}$ larger than 0.11 and 0.30, respectively) were not considered in the subsequent analysis. As we were interested in a comparison with other galaxies such as the Magellanic Clouds, we considered only the stars with $P \le 30$ d, since for longer P the number of Cepheids per galaxy with available light curves is rather small.

3 Results

The Fourier parameters of the Cepheids in IC 1613 and NGC 6822 are compared in Fig. 1 with those of the stars in the Milky Way (MW) and the Magellanic Clouds (MCs). The data on the MW stars were taken from our previous works (e.g. Antonello & Morelli 1996), while the data on the MCs stars were taken from Udalski et al. (1999a,b). Unlike for the MW, the best photometry of the MCs Cepheids was obtained by OGLE in the I band, so we adopted the two different bands V and I for the MW and MCs, respectively. This adoption does not allow direct comparison of the parameters of the Cepheids of the MW with those of the MCs. In general the R₂₁ values are not very sensitive to the filter in the optical range (Simon & Moffett 1985). On the other hand, we estimated a systematic difference $\phi_{21}(I) - \phi_{21}(V) \sim 0.33$ between I and V photometric bands. This estimate was made using OGLE data of some MCs Cepheids with good V light curves.

The plots in Fig. 1 show several interesting features.

(1) From the panels (a) and (b) it is evident that the Cepheids with P close to 10 d of IC 1613 and NGC 6822 do not show the high $\phi_{21}$ values ( $\ga$ 5.5 rad) of MW and MCs Cepheids; the result was already known for IC 1613 and is confirmed for NGC 6822. The obvious question is whether this is only due to the poor number of Cepheids in IC 1613 and NGC 6822. If we consider the $\log P$ range between 0.9 and 1.2, there are 127, 82, and 14 stars in the MCs, MW, and IC 1613 plus NGC 6822, respectively, whereas there are 35 MCs stars with $\phi_{21}(I) \ga 5.7$ in this period range and 13 MW stars with $\phi_{21}(V) \ga 5.4$ , that is, about 28% (MCs), and 16% (MW), respectively. These statistics are poor; however, one would also expect about three stars with a high value of $\phi_{21}$ also in IC 1613 and NGC 6822. We do not think that the lack of such stars is due to a selection effect related to a possibly very small amplitude. For example, in IC 1613 both Antonello et al. (1999) and Udalski et al. (2001) detected the Cepheid V2414-A = OGLE13808 with P=7.58 d and amplitude of about 0.2 mag; moreover, only part of the stars with high $\phi_{21}$ values have a relatively small amplitude. Therefore we think that the lack of Cepheids with large $\phi_{21}$ values in IC 1613 and NGC 6822 is probably due to some physical reason related to the pulsational characteristics.

(2) In panels (c) and (d) there are no evident differences among the R₂₁ values in the various galaxies for $1 \la \log P \la 1.48$ . It is known that the Cepheids with $\log P \la 1$ of SMC have progressively higher R₂₁ values than do LMC ones. The distribution of Cepheids with $\log P \la 1$ of IC 1613, and in particular those of NGC 6822, is shifted with respect to that of MW Cepheids; a slight shift is possible even with respect to the MCs ones. In order to estimate the possible statistical significance of the shifts, a linear fit was applied to the Cepheids in the range $0.7 < \log P < 1.05$ , with average $\log P$ value 0.875. In the panel (d), the MW Cepheids give

$\begin{displaymath}R_{21}(V)=0.260(\pm.005)-0.963(\pm.048)(\log P - 0.875). \end{displaymath}$

(1)

The mean separation of the Cepheids of NGC 6822 and IC 1613 from this line is 0.084 with $\sigma=\pm.013$ ; that is, the significance of the shift is larger than 6 $\sigma$ . In the panel (c) the shift with respect to the linear fit of MCs Cepheids

$\begin{displaymath}R_{21}(I)=0.287(\pm.004)-0.831(\pm.041)(\log P - 0.875) \end{displaymath}$

(2)

is marginal, $0.035\pm.014$ . One could suspect in particular that in the narrower range $0.90 \la \log P \la 1.05$ the stars of NGC 6822 and IC 1613 have higher R₂₁ values than those of MW and MCs.

(3) The distribution of the stars with higher $\phi_{21}$ values in panel (a) shows a different trend from that in panel (b). The different trend is not explained by the different photometric band, and its reality, in spite of the poor number of Cepheids, can also be verified in panels (e) and (f), which only show the stars in the range $0.9 \la \log P \la 1.2$ . We have drawn a dashed line by eye to mark the trends. In this context we note that there are no evident differences between the Cepheids of the two Clouds, hence we have not used different symbols for SMC and LMC stars in the plot.

4 Discussion and conclusion

This comparison of the Fourier parameters of Cepheid light curves for different galaxies of the Local Group indicates some possible differences between MW, MCs, IC 1613, and NGC 6822. The physical reason, however, is not completely clear, because such differences do not appear to be a simple metallicity effect. In fact, though IC 1613 is the metal poorest among the four galaxies, and the metallicity of NGC 6822 is slightly larger than that of SMC (Venn et al. 2001), NGC 6822 shows the largest differences with respect to the other galaxies.

Radiative models predict a sensitivity of the Fourier parameters of the light curves to the metal content for P₀ close to the resonance center P₀/P₂=2(e.g. Buchler 1998), but this high sensitivity is not confirmed by the observations of Cepheids in the MCs compared to those in the MW; moreover, it seems that including convective transport and turbulent dissipation is not sufficient for improvements. We recall in conclusion that it seems that pulsational models are not yet able to satisfactorily reproduce the light curve features observed in the different galaxies, so we hope that the present results will supply further useful observational indications for comparison with model predictions.

References