Astron. Astrophys. 349, 55-69 (1999)

5. Results

About 110 stars were detected as variables: population I and II Cepheids, eclipsing binaries, long period, semiregular and irregular variables. In Table 3 we have reported the stars discovered by Baade and by Sandage (Sandage 1971) falling in field A, their period, variable type, the identification number in the present work and the period obtained by us. All the previously known Cepheids are confirmed; the periods obtained with our data are very similar to the old ones. No attempt has been done for increasing the significant digits of the periods by analysing old and new data together. For the irregular variables we have reported the time scale of variability which is compatible with our data. Some of the stars are discussed in the following subsections; in particular, the unusual characteristics of the star SV39 (V1740) are also confirmed, and they are discussed in Sect. 5.2.

Table 3. Variable stars in common with Baade-Sandage

Cepheids, other periodic variables, eclipsing binaries and irregular or semiregular variables are listed in Tables 4, 5, 6 and 7, respectively. The stars are identified by their name; the right ascension and declination are given along with the P and the mean Wh magnitude (for eclipsing binaries an estimate of Wh at the maximum luminosity is reported). For Cepheids the probable pulsation mode is also reported, while for the irregular, possible long period and semiregular variables (Table 7) the timescale of variability is indicated. The astrometric positions were computed using 11 previously known variable stars as local astrometric standards, to derive transformation equations from the CCD positions to (1950) and (1950). The coordinatae were taken from the General Catalogue of Variable Stars (GCVS; Samus 1995). Other four stars, SV36 (V1756), SV40 (V1872), SV52 (V1908) and SV43 (V2321), were excluded because the values in GCVS differ by some arcsecs from the results of the transformation equations; this indicates an identification problem. The comparison of the derived positions with those reported by Freedman (1988b) shows that the accuracy of the transformation is generally better than about 0:005.

Table 4. Cepheids in field A of IC1613

Table 5. Other periodic variables in field A of IC1613

Table 6. Eclipsing binaries in field A of IC1613

Table 7. Irregular and possible long period and semiregular variables in field A of IC1613

The variable stars with known index are shown in the color-magnitude diagram of Fig. 7.

Fig. 7. Color-magnitude (V-R,V) diagram. Upper panel: open circles: population I Cepheids of fundamental mode; filled circles: population I Cepheids of first overtone mode; open triangles: eclipsing binaries. Lower panel: filled triangles: periodic variables (those with V-R less than about 0.6 are population II Cepheids); asterisks: other semiregular and irregular variables

5.1. Cepheids

The light curves of the detected Cepheids are displayed in Figs. 8 and 9; note that the magnitude scale is not the same in the different panels. Just from a simple inspection it is possible to conclude that both fundamental and first overtone mode Cepheids have been detected; the fundamental mode Cepheids have large amplitude or asymmetric light curves, while first overtone modes have relatively small amplitude and more symmetric light curves. We have used the Fourier parameters and the amplitudes for discriminating the pulsation mode. A detailed discussion of these parameters and comparisons with other galaxies will be reported in Paper II (Antonello et al. 1999). There is no reliable indication of double-mode Cepheids; probably the precision and the sampling of the data are not sufficient for their detection.

Fig. 8. Cepheid Wh light curves. For each star, the identification number and the period are reported. Note the different magnitude scales. The V-Wh range of Cepheids is 0.1-0.4 mag, and therefore for these stars V is larger than Wh by some dex

Fig. 9. Cepheid Wh light curves (see Fig. 8)

The stars with known V-R occupy a vertical band in the color-magnitude diagram (Fig. 7), or instability strip. The PL diagram for the Wh-band is briefly discussed in Sect. 6.2 and shown in Fig. 15.

In the following we report some notes on selected stars.

V2396 . The data sampling does not allow to construct the complete light curve, however the period found by us is close to that given by Sandage (1971).

V2414 and V0107 have very small amplitudes, about 0.2-0.3 mag; looking at our Galaxy, this is not unusual for stars with P between 7 and 10 d.

V1337 has the best light curve, which can be fitted with a 5th order Fourier decomposition and rms residual of 0.026 mag.

V2221 is an interesting case, because it has a close (0:008) companion which is slightly fainter. DAOPHOT was not able to resolve always the two stars, and therefore the resulting light curve was very scattered. We have simply summed the luminosities of the two stars when they were resolved; the two light curves are shown in Fig. 10. Some scatter is still present, but the Cepheid behavior is evident; clearly its amplitude is smaller than what should be expected, and one should correct for the companion's luminosity before using the star in a PL relation. The referee has remarked, however, that these problems given by close stars can be overcome by using the fixed position photometry (see Kaluzny et al. 1998).

Fig. 10. Wh light curve of the Cepheid V2221 showing the effect of a close star (about 0:008). Upper panel: in many cases the two stars are not discriminated by DAOPHOT; lower panel: when resolved, the luminosities of the two stars are summed in order to get a consistent light curve

V1592 was not noticed by Baade and Sandage, even if it has large amplitude and is sufficiently bright, probably because it is located in a partially crowded region.

V0551, V0655 and V2100 are characterized by rather symmetric light curves, but their periods and luminosities are typical of fundamental mode Cepheids; the available color of one of them indicates a location in the instability strip. The nature of these stars is uncertain; tentatively we put them in relation with the anomalous Cepheids, even if their periods and luminosities are larger than those seen in galactic and extragalctic anomalous Cepheids.

5.2. SV39=V1740

The enigmatic nature of this variable was already pointed out by Sandage (1971). According to the old data the light curve could be described as an inverted Lyrae eclipsing variable with a period of 28.72 d. Notwithstanding this very peculiar shape and the high luminosity with respect to the PL relation, Sandage included it among Cepheids and as such it remains in the GCVS. We have reanalyzed by means of the power spectrum technique both the old photograpic measurements (107 datapoints) and our 67 Wh ones. The analysis was performed both separately for the two data sets and by merging the sets after rescaling the Wh data to the B ones (the scale factor is 1.9) and aligning the zeropoints of the two timeseries. The light variation seems rather complicated but a firm conclusion can be reached: there are at least two periodic terms which fit both datasets, one with a period of 28.699 d, very similar to that suggested by Sandage, and the other with P=1123 d. The reality of this long period term can be deduced from Fig. 11. In the upper panel we have plotted all the data phased with the short period. The photographic data are indicated by asterisks, while filled circles represent the Wh data of the season 1995-97, and open circles those of 1998. It is evident that in 1998 the star was systematically less bright than in 1995-97. In the lower panel we have subtracted from the data the long period term. Now the data dispersion about a mean curve of both photographic and Wh data is considerably decreased, and in particular the 1998 Wh data are well aligned with those of previous years. Another fact can be deduced from the figure: after removing the long period term, the difference between the two maxima is significantly decreased. As a matter of fact if we adopt as the short period 14.350 d instead of its double, the fit of the data is only marginally worse. Therefore with the present available dataset, we prefer to be cautious and not do decide which of the two possible short periods is the correct one. Another and more pregnant open question is the nature of the variability. The color of the star, V-R=0.52, is similar to that of Cepheids; however no radially pulsating star with such an amplitude is known to have symmetric maxima, and furthermore it is not possible to explain theoretically such a shape. A check of the amplitudes is not much conclusive: assuming the amplitude was essentially constant during fifty years, the ratio of photographic and white light amplitudes is in the range 0.9-1.5 for 5 bright Cepheids, and only V1039 has ratio 1.9 as V1740.

Fig. 11. Light curve of V1740 (SV39). Filled circles: 1995-1997 Wh data; open circles: 1998 Wh data; asterisks: photographic B data published by Sandage (1971); the Wh data have been rescaled (see text). Upper panel: data phased with d; lower panel: data phased with the same period after subtracting the component with  d

5.3. Other periodic variables

Periodic variables include red variables and population II pulsating stars (Fig. 12), and eclipsing binaries (Fig. 13).

Fig. 12. Wh light curves of periodic variables.

Fig. 13. Wh light curves of eclipsing binaries.

V1892, V1193 and V1978 are probable long period variables. In particular, V1193 could be an RV Tau star with a long period of 160.2 d; however V-R is 0.94.

The population II Cepheids (or W Vir stars) should have V-R not very different from that of population I Cepheids, since the latter are rather metal-poor; moreover, for a given luminosity, population II Cepheids have a much longer period. On these basis we have identified 5 of such stars, namely V00130, V0881, V0971, V1598 and V1935 .

From a simple inspection of the phased light curves we have identified 24 possible eclipsing binaries. Only 8 stars are reported in Table 6 and are shown in Fig. 13. A detailed anlaysis of all the candidates will be performed in a subsequent paper.

5.4. Other semiregular and irregular variables

There are several stars which are characterized by irregular variability on different time scales; some of them could be long period variables which cannot be identified as such owing to the data sampling and the short observing time interval. In Fig. 14 some of these stars are shown.

Fig. 14. Wh light curves of a sample of irregular variables; note the different magnitude scales

SV21=V3106 could vary both with short (tens of days) and long ( d) timescales; the amplitude is about 1 mag and the color is V-.

SV52=V1908 is variable with small amplitude (less than 0.3 mag), and appears brighter than fifty years ago. The color V-R=0.39 indicates a yellow star in the upper part of the instability strip.

If the identification is correct, the star SV56=V1800 previously known as irregular variable, appears to be constant or variable with very small amplitude (less than about 0.1 mag); it is not reported in Table 7.

© European Southern Observatory (ESO) 1999

Online publication: August 25, 1999
helpdesk.link@springer.de