当前位置:首页 >> >>

Time Series Photometry of Variable Stars in the Globular Cluster NGC 6397

Time Series Photometry of Variable Stars in the Globular Cluster NGC 6397
J. Kaluzny1 , and I. B. Thompson2 ABSTRACT Time series BV I photometry is presented for 16 short-period variables located in the central region of the globular cluster NGC 6397. The sample includes 9 newly detected variables. The light curve of cataclysmic variable CV6 shows variability with a period of 0.2356 days. We con?rm an earlier reported period of 0.472 days for cataclysmic variable CV1. Phased light curves of both CVs exhibit sine-like light curves, with two minima occurring during each orbital cycle. The secondary component of CV1 has a low average density of 0.83 g cm?3 indicating that it cannot be a normal main sequence star. Variables among the cluster blue stragglers include a likely detached eclipsing binary with orbital period of 0.787 days, three new SX Phe stars (one of which has the extremely short period of 0.0215 days), and three low amplitude variables which are possible γ Doradus variables. Subject headings: binaries: close - globular clusters: individual (NGC 6397) novae - cataclysmic variables - blue stragglers - stars:variables:other

arXiv:astro-ph/0210626v1 29 Oct 2002



The inner region of the nearby post-core-collapse globular cluster NGC 6397 has been the target of several surveys aimed at the identi?cation of rare or unusual stars likely to be created as the result of stellar interactions in its dense cluster core (Coll & Bolton (2002) and references therein). Recently Grindlay et al. (2001) reported the detection with Chandra of 25 X-ray sources within 2′ of the cluster center. Optical searches for variable stars in the inner regions of globular clusters, particularly those with strong central density cusps, have been hindered by crowding. Image subtraction techniques in such strongly crowded ?elds have been successful in locating and studying variables (Olech et al. 1999; Kaluzny, Olech
1 2

Copernicus Astronomical Center, Bartycka 18, 00-716 Warsaw, Poland; jka@camk.edu.pl Carnegie Observatories, 813 Santa Barbara St., Pasadena, CA 91101-1292; ian@ociw.edu

–2– & Stanek 2001), especially when the data are taken with a ?ne plate scale with a stable and uniform point-spread function. In this paper we present the analysis of ground-based time series CCD photometry which was obtained to study the light curves of the optical counterpart to the binary millisecond pulsar PSR J1740-5340 (D’Amico et al. 2001b; Ferraro et al. 2001). Our results for the binary pulsar are given elsewhere (Kaluzny et al. 2002). In this contribution we report on the results obtained for other variable stars located in the central part of the cluster. Light curves for a total of 16 variables are presented and discussed. Nine of these objects are new identi?cations.



The photometric data were obtained with the 2.5-m du Pont telescope at Las Campanas Observatory. A ?eld of 8.65 × 2.60 arcmin2 was observed with the TEK#5 CCD camera at a scale of 0.259′′ /pixel. This present analysis is limited to a sub-?eld 2.60 × 2.60 arcmin2 centered approximately on the cluster core. Most of the data were obtained on 6 nights during the period from UT May 1 to 8, 2002, with additional data obtained on UT June 3, 2002. Conditions were non-photometric on all but one night with average seeing of 1.0′′ in the V -band. The cluster was observed for a total of 32 hours through BV IC ?lters. Exposure times were typically 30 sec (B), 15 sec (V ), and 10 sec (I). Frames were were co-added, and the total number of stacked images used in the present study was 69 (B), 196 (V ), and 59 (I). For a few of the variables discussed below we have also extracted V -band timeseries photometry from images with the best seeing, selecting 1176 out of the total of 1256 available. Our observing material and photometric calibration procedure are described in detail in Kaluzny et al. (2002).



Two methods were used to detect potential variable stars. Both of them make use of the ISIS-1.2 image subtraction package (Alard & Lupton 1998; Alard 2000). The ?rst method is based on examination of images created by combining individual residual images, and relies entirely on tools included in the ISIS package. This method is

–3– well suited for the detection of variables with a high duty cycle 3 and/or showing signi?cant changes of ?ux (bright variables or faint variables with large amplitude light curves). An advantage of this method is that it permits the detection of variables which cannot be resolved with classical pro?le photometry in crowded ?elds. The second method relies on the examination of the light curves of all objects which could be measured on a reference frame with pro?le ?tting software. For each ?lter a reference frame was constructed by averaging several stacked images of the best quality. The detection of stellar objects and the extraction of photometry was accomplished using the DAOPHOT/ALLSTAR software package (Stetson 1987). The total of 4336 stars were measured on the V -band reference image. The limiting magnitude depends very much on the distance from the cluster core. The faintest measured stars have V ≈ 21.3 and the observed luminosity function starts to diminish at V ≈ 19.0. The phot procedure in the ISIS package was used to extract di?erential light curves at the position corresponding to each star detected with the DAOPHOT/ALLSTAR package. Di?erential light curves were then transformed to magnitudes and checked for variability. The light curves were searched for the presence of any periodic signal with the AoV algorithm (Schwarzenberg-Czerny 1989; Schwarzenberg-Czerny 1996) and were reviewed for possible eclipse-like events. A total of 16 variables were identi?ed. Nine of these are new detections. Information on these variables is presented in Table 1. Column 1 gives an assigned name following Clement et al. (2001), followed by the right ascension and declination of the variable 4 . Column 5 gives alternate designations for the variables, many of which have been previously identi?ed as blue stragglers, UV bright objects, and/or candidate cataclysmic variables. Table 1 also gives positional information for all 16 variables. Columns 2 and 3 list equatorial coordinates, while column 4 lists pixel coordinates of the variables as found on HST archive image u5dr0401r for 12 of the 16 variables. The equatorial coordinates were derived from the frame solution included in the HST image header. We derived equatorial coordinates for the four variables laying outside the HST ?eld from an astrometric solution to our V band reference image based on 1022 stars with equatorial coordinates adopted from Kaluzny (1997). This solution has residuals not exceeding 1′′ when compared to the HST astrometric solution. Finding charts for variables V10, V22 and V23 can be found in
By ”duty cycle” we mean the fraction of time in which a given star shows luminosity other than its median luminosity. An example of an object with low duty cycle is an eclipsing binary with narrow eclipses. Clement et al. (2001) list 12 variables in NGC 6397, V1-V12. Four variables listed in Table 1 are present in the Clement et al. paper, and we copy their name assignments (V7, V10, V11 and V12). Variables V13-V24 are not in the Clement et al. catalogue.
4 3

–4– Lauzeral et al. (1992; stars 11, 8 and 16 in their Fig. 1), and a ?nding chart for variable V24 is shown in Figure 1. Basic information on the photometric properties of the 16 detected variables is presented in Table 2. Column 1 lists the variable name, followed by a classi?cation of the light curve, the period of variability, mean B, V , and I magnitudes, and the V -band full amplitude of variability. Positions of the variables in the cluster color-magnitude diagram are shown in Fig. 2. The ellipsoidal variable V16 is the optical counterpart to the millisecond binary pulsar J1740-5340 (D’Admico et al. 2001; Ferraro et al. 2001). Photometry of this variable is discussed in Kaluzny et al. (2002).


ANALYSIS OF PHOTOMETRY 4.1. Cataclysmic Variables

Grindlay et al. (2001) detected 9 possible cataclysmic variables (CV) in the central region of NGC 6397 with the Chandra telescope, naming the objects CV1 - CV9. They identi?ed the optical counterparts of CV1 - CV5 based on observations reported by Cool et al. (1995, 1998), Grindlay (1999), and unpublished HST H-alpha observations (Grindlay et al. 2001). Variable V12 can be unambiguously identi?ed with CV1 using positional data provided in Cool et al. (1998; see their Table 1). From an examination of the HST archive image u5dr0401r at the equatorial coordinates of V13 we conclude that this star is the optical counterpart of CV6. On the HST image V13 is visible as an isolated object with a closest neighbor at a distance of d ≈ 0.8′′ . Variables CV2 - CV5 and CV7 - CV9 could not be resolved on our reference images. We attempted to extract di?erential light curves using the ISIS package at the known positions of these stars. The light curves su?ered from large photometric errors and showed no sign of any periodicity. The large errors result from the e?ects of several relatively bright stars near the variables.

4.1.1. CV1 Examination of the light curves of V12 = CV1 reveals the presence of a sine-like periodic modulation. The power spectrum calculated from the V ?lter time series photometry based on individual exposures is presented in Fig. 3. Two maxima of comparable strength are

–5– present at frequencies corresponding to periods of 0.4712 and 0.2356 days. Grindlay et al. (2001) report that CV1 showed one total eclipse in X-rays through the 0.567 days observation obtained with Chandra, and therefore we may eliminate the shorter period from consideration. Nightly light curves of CV1 phased with the period of 0.4712 days are displayed in Fig. 4. The average value of the formal error of a single data point is 0.042 mag. It is worth noting that the shape of the light curve as well as the average luminosity of the system are relatively stable over the interval covered by our observations. Figure 5 shows phased BV I light curves corresponding to photometry extracted from averaged images. Neither the amplitude nor the shape of the light curves change noticeably with band-pass. The light curves are quite symmetric with two maxima of comparable height separated by half of the period. The two observed minima have comparable depths although the minimum occurring at phase zero is slightly sharper. These properties of the light curves of CV1 suggest that the observed variability is dominated in optical domain by the ellipsoidality e?ect caused by rotation of the Roche lobe ?lling component, and that the 5 secondary dominates optical ?ux of the system. Such an interpretation is consistent with the relatively red colors of CV1. The variable is located about 0.1 mag to the blue of the cluster main sequence on the V /B ?V plane and it is located slightly to the red of the main sequence on the V /V ?I plane (See Fig. 2). At the time of our observations the optical ?ux generated by the accretion process apparently contributed a small fraction of the total optical luminosity of the binary. Another interesting property of CV1 is its relatively long orbital period. Among 318 cataclysmic variables which are listed in Ritter & Kolb (1998) there are only 14 objects with periods longer than 0.47 days. It is possible to get a robust and reliable estimate of the average density of the secondary component of CV1 from the formula in Faulkner et al. (1972) and Eggleton (1983): < ρ >= 107P ?2 (1)

where the period P is in hours and the density ρ is in g cm?3 . We obtain < ρ >= 0.83 g cm?3 . The binary is located well below the cluster turn-o? and therefore we may expect that mass of the secondary does not exceed 0.8M⊙ . Theoretical models published recently by Bergbusch & VandenBerg (2001) give an average density < ρ >= 3.52 g cm?3 , for a ZAMS model of a 0.8 solar mass star with [Fe/H] = ?2.0. For a mass lower than 0.8 m⊙ the expected density is even higher as < ρ >? m?2 . We conclude that the secondary component of CV1 is noticeably over-sized compared to a normal low-mass main-sequence star. Knowing the distance modulus of the cluster we may derive an absolute luminosity of
For cataclysmic variables it is conventional to call the degenerate component the ”primary” and its companion the ”secondary”, independent of the actual mass ratio of a given system.

–6– the variable. For (m ? M)V = 12.69 ± 0.15 and E(B ? V ) = 0.18 (Reid & Gizis 1998) we obtain < MV >= 4.7 for the average absolute magnitude of CV1 in the V -band. It is tempting to use that information to derive the radius of the secondary star but we feel that uncertainties in relative intensity of the accretion generated ?ux to the total observed luminosity are too large. Such uncertainties a?ect not only the estimate of the ?ux from the secondary but also any estimate of its e?ective temperature. These problems can be greatly reduced by observing the binary at near-IR wavelengths where the total ?ux of the system should be strongly dominated by the secondary star. We conclude this part of the discussion by noting that X-ray observations presented by Grindlay et al. (2001) are consistent with the identi?cation of CV1 as either an ordinary dwarf nova or a magnetic CV.

4.1.2. CV6 The power spectrum of the V band light curve of V13 = CV6 shows strong peaks at periods of 0.1176 and 0.2352 days. An examination of the light curves from individual nights indicates that they exhibit two minima of di?erent shape separated by about 0.12 days. This is particularly clear in the light curve extracted from individual images collected on the night of UT May 1, 2002, which is presented in Fig. 6. The BV I light curves of CV6 phased with a period of 0.2352 days are shown in Fig. 7. They are based on photometry extracted from averaged images. The shape and mean level of the light curves were quite stable during our observations. As for CV1 the variability of CV6 seems to be dominated by the ellipsoidality e?ect. The minimum occurring at phase 0.0 is narrower than the minimum at phase 0.5. We interpret this as evidence that the bright accretion region surrounding the primary component of the binary is eclipsed at phase 0.0. The variable is located about 0.15 mag to the blue of the cluster main sequence in the V /B ? V plane (see Fig. 2) and in the V /V ? I plane it is located at the red edge of the cluster main sequence. Assuming cluster membership for CV6 we estimate MV = 6.2. The average density of the secondary component is ? 3.4 g cm?3 , consistent with the density expected for a slightly evolved Pop II main sequence star of mass 0.7 m⊙ . In particular, models published by Girardi et al. (2000) predict < ρ >= 3.2 g cm?3 and MV = 6.2 for 0.7 solar mass star of age 11 Gyr.

–7– 4.2. Eclipsing Binaries

Our sample of variables includes four eclipsing binaries. In this section we comment brie?y on their properties, a detailed analysis will require spectroscopic observations. Variable V7 was identi?ed as a W UMa variable by Kaluzny (1997). This star has two close visual companions with V = 18.31 and V = 19.07 located at angular distances of 0.54′′ and 0.56′′ , respectively. Despite the proximity of the companions the pixel scale of the observations together with the high signal-to-noise data meant that photometry could be measured for both companions in our V band data. Only the brighter companion could be measured while extracting photometry for the B and I bands. Both companions were unresolved in the photometry reported by Kaluzny (1997), leading to an overestimation of the luminosity of V7 as reported in that paper. Variable V19=PC-1 was identi?ed by Taylor et al. (2001) in a photometric survey for objects with excess Hα ?ux. They also detected V7 in the course of their survey. V19 has a close visual companion of V = 17.22 at angular distance of 0.5′′ . That companion is measured in our photometry for all bands. Phased V -band light curves of V7 and V19 are shown in Fig. 8. Some intrinsic nightto-night changes of the shape of the light curve were observed for V7. Such behavior is not unusual for W UMa type systems. There is some indication that the secondary minimum of V7 exhibits a ”?at-bottom” indicating that this eclipse is total. For W UMa type systems with total eclipses one may obtain reliable light curve solutions as totality removes the degeneracy between the mass ratio and inclination of the system (Mochnacki and Doughty 1972). These two contact binaries have similar colors and lie about 0.7 mag above cluster main sequence on the color-magnitude diagram, suggesting that that they are most likely members of NGC 6397. Variable V14 is a relatively faint eclipsing binary located about 0.1 mag to the red of the cluster main sequence. Its phased light curve is presented in Fig. 9, two shallow minima of di?erent depth are seen. Main sequence binaries with periods below 0.35 days almost always show ordinary EW type light curves, a signature of a contact con?guration. If V14 is a member of NGC6397 then its red color and faint magnitude would suggest the components are of late spectral type with low masses and radii. In this case V14 could be a close but non-contact binary despite its short period. Variable V18 is by far the most interesting of the four eclipsing binaries included in our sample. It is not only a likely blue straggler but it also shows a very unusual light curve (see Fig. 10). At ?rst glance it resembles the light curves of ordinary W UMa type contact binaries. However the light curve of V18 shows clear signatures of eclipse ingress and egress, not observed in contact binaries. We conclude that V18 is a detached or semi-detached

–8– system composed of stars with very similar surface brightness. The light curve is similar in all 3 ?lters with some indication that eclipses become progressively shallower from the B to the I band by ? 0.02 mag. Examination of HST images of the cluster shows that V18 possesses a close visual companion at an angular distance of 0.19 ′′ . It has not been resolved in our pro?le photometry and hence its ?ux acts as a ”third” light in the photometry of V18. We have identi?ed the companion as star 200338 in the data base published by Piotto et al. (2002), with V = 18.631 and B = 19.338. We adjust these magnitudes to V = 18.613 and B = 19.281 to take into account di?erences in the zero points of the two sets of photometry of ?0.018 and ?0.057 for the V and B ?lters, respectively (our magnitudes are brighter). Since I-band data are not included in the Piotto et al. (2002) study we estimate I-band photometry from the fact that the V18 companion lies on the cluster main sequence. From the V I photometry of NGC 6397 published by Alcaino et al. (1997) we estimate that for V = 18.61 the companion has I ≈ 17.52. Light curves presented in Fig. 10 as well as magnitudes listed for V18 in Table 2 are corrected for the contribution of the nearby companion. Attempts to derive a reliable light curve solution for V18 are hampered by the lack of vital information on the mass ratio of the binary. We have calculated a grid of solutions for a wide range of assumed values of the mass ratio q = m2 /m1 . Index ”1” refers to star eclipsed at phase ”0”. The V -band light curve was was solved using Wilson-Devinney code (Wilson 1979) embedded in the MINGA minimizing package (Plewa 1988). Our preliminary results can be summarized as follows. For 0.08 < q < 2.35 solutions imply a detached con?guration with an inclination in the range 61 < i < 74 deg. The ratio of component radii is constrained to the range 0.83 < r2 /r1 < 1.17. For q > 2.35 solutions converge to semi-detached con?gurations with the less massive component ?lling its Roche lobe. The χ2 statistic measuring the quality of ?t of the synthetic light curve to the observations has a minimum near q = 0.20. A solution for that speci?c value of the mass ratio gives inclination i = 71 deg and average relative radii of the components r1 = 0.22 and r2 = 0.24. If the mass ratio is close to 0.2 and the system is detached, then one may wonder why both components have very similar e?ective temperatures. Note that the color of the variable is essentially constant over the whole orbital period. However, if the mass ratio is close to unity then we face di?culty trying to explain how this blue straggler can be composed of two stars both of which are signi?cantly bluer and more luminous than stars at the cluster turno?. Spectroscopic data providing information about mass ratio of the binary are needed to reliably determine its geometrical and absolute parameters.

–9– 4.3. SX Phe Stars

SX Phe type variables are short period pulsating stars which can be considered Pop II counterparts of more metal rich δ Sct type stars. It is not unusual to ?nd them among blue stragglers in globular clusters (Rodriquez & L?pez-Gonz?les 2000). Variables V10 and V11 o a were originally identi?ed as SX Phe stars by Kaluzny (1997). Here we add three more objects to that group. Light curves of all ?ve variables show modulation of shape and amplitude indicating the presence of multimodal pulsations. A detailed analysis of these data will be published in a separate paper (Schwarzenberg-Czerny et al.; in preparation). Here we note only that variables V10 and V15 have extremely short dominant periods, at 0.0215 days V15 has the shortest period known for an SX Phe star. No SX Phe stars with periods below 0.030 days are listed in the recently published catalog of Rodriquez & L?pez-Gonz?les (2000). We o a considered the possibility that V15 is not an SX Phe type star but rather a pulsating hot subdwarf. However its B ? V color is too red to be a bright sdB/O star (note the position of V15 in Fig. 2).


Other Variables

In this section we discuss brie?y four variables which cannot be classi?ed with con?dence based on the available data. Variable V20 is one of the brightest blue stragglers identi?ed in the cluster core by Lauzeral et al. (1993). The power spectrum of its light curve shows two major peaks at periods of 0.861 and 0.436 days, with the higher peak corresponding to the longer period. The V band light curves of V20 phased with each of these two periods are shown in Fig. 11. For the longer period the light curve has two minima, and this suggests that V20 is a low amplitude W UMa type system. The feature visible at the second quadrature arises from a di?erent light level observed on the single night of June 3, roughly one month after the ?rst observing run when most of the data were collected. The period of 0.861 days is relatively long for a contact binary belonging to a globular cluster (Rucinski 2000). Yet another possibility is that V20 is a close binary with variability due to the ellipsoidality e?ect. The light curve phased on the shorter period is slightly more noisy with a single minimum. The period of 0.436 days is far too long to classify V20 as a pulsating SX Phe star. The variable is too hot to show spot-related activity as is observed for FK Com or BY Dra type stars. However, it can be related to γ Doradus stars, as it is discussed below for two other variables. Phased V -band light curves for variables V17 and V24 are presented in Fig. 12. They

– 10 – both show low amplitude, sine-like modulation, with periods of 0.457 days and 0.525 days for V24 and V17, respectively. On the color-magnitude diagram the stars are located about 0.15 mag to the red of the cluster turno?. We propose that V17 and V24 are Pop II counterparts of γ Doradus variables. The γ Doradus stars have often multiple periods between 0.4 and 3 days and show sinusoidal light curves with amplitudes – in optical domain – of the order of 0.01 mag (Zerbi 2000; Henry & Fekel 2002). Their variability is due to non-radial gmode pulsations and they are usually subgiants or, less frequently, main sequence stars of spectral type F0-F2. The red edge of the instability strip for Pop I γ Doradus variables is located at (B ? V ) ≈ 0.37 (Henry & Fekel 2002). The dereddened colors of V17 and V24 are (B ? V )0 = 0.45 and (B ? V )0 = 0.37, respectively. Henry & Fekel (2002) note that γ Doradus variables show average A(B)/A(V ) amplitude ratios of ? 1.3. This distinguishes them from ”spotted” variables which have amplitude ratios of ? 1.1 and from ellipsoidal variables for which the ratio is ? 1.0. From our data we obtain A(B)/A(V ) = 1.09 ± 0.21 and A(B)/A(V ) = 1.25 ± 0.14 for V17 and V24, respectively. More extended time series would allow a more accurate estimate of the A(B)/A(V ) ratio for both variables and would also help to search for multiperiodicity in the light curves. The stability of γ Doradus light curves 100-200 cycles also distinguishes these stars from ”spotted” variables. Variable V22=BS8 is a bright blue straggler which was identi?ed in the cluster core by Lauzeral et al. (1993). The light curve of V22 cannot be phased with a single period although during May run we observed four minima occurring in 1 and 2 day intervals. The light curve extracted from observations obtained on nights of 9 and 10 July, 1995 (Kaluzny 1997) shows variations with δV ≈ 0.12 and a possible period of about 0.75 days. However that period does not ?t the 2002 data. It is possible that V22 is a pulsating multiperiodic variable related to γ Doradus stars or that it is a distant RR Lyr variable of RRd type. In Fig. 13 we show light curves from two nights on which the observed variations were most pronounced.



We have used time series photometry obtained with a medium sized telescope to look for short period variables in the central part of the post-core-collapse cluster NGC 6397. We show that by applying the image subtraction technique that it is possible not only to detect variable stars in very crowded ?elds but also to measure accurate light curves for objects with amplitudes as small as 0.01 mag. Photometry of NGC 6397 obtained with HST imaging (Piotto et al. 2002) allows a check and, if necessary, a correction for contamination from possible visual companions which are unresolved in ground-based data.

– 11 – We present the ?rst complete light curves and derive orbital periods for two cataclysmic variables in NGC 6397. The total ?ux and variability of both of these CV’s is dominated by the secondary components. Several new variables have been identi?ed among the cluster blue stragglers, including one detached binary and three objects being possible Pop II counterparts of γ Doradus stars. The determination of cluster membership of the detected variables has relied on their positions in the cluster CMD. While the cluster is located on the sky near the Galactic bulge at l = 338 deg and b = ?12 deg, in the central region cluster stars must prevail strongly. However we cannot exclude the possibility that some of the variables are ?eld interlopers. The publication of proper motion catalogs (Cool & Bolton 2002) will be exceptionally useful in resolving issues of cluster membership. JK was supported by the Polish KBN grant 5P03D004.21 and by NSF grant AST9819787. IT was supported by NSF grant AST-9819786. We would like to thank Alex Schwarzenberg-Czerny for providing us with his excellent period ?nding programs.

REFERENCES Alard, C. 2000, A&A, 144, 363 Alard, C., & Lupton, R. H. 1998, ApJ, 503, 325 Alcaino, G., Liller, W., Alvarado, F., Kravtsov, F., Ipatov, A., Samus, N., & Smirnov, O. 1997, AJ, 114, 1067 Bergbusch, P.A., & VandenBerg, D.A. 2001, ApJ, 556, 322 Clement, C. M. et al. 2001, AJ, 122, 2587. Cool, A.M., Grindlay, J.E., Cohn, H.N., Lugger, P.M., Bailyn, C. 1998, ApJ, 508, L75 Cool, A.M., & Bolton, A.S. 2002, in ”Stellar Collisions, Mergers and their Consequences”, ASP Conf. Series, Vol. 263, ed. M.Shara, in press (astro-ph/0201166) D’Amico, N., Possenti, A., Manchester, R. N., Sarkissian, J., Lyne, A. G., & Camilo, F. 2001b, ApJ, 561, L89 Eggleton, P.P. 1983, ApJ, 268, 386

– 12 – Faulkner, J., Flannery, B.P., & Warner, B. 1972, ApJ, 175, L79 Ferraro, F. R., Possenti, A., D’Amico, N., & Sabbi, E. 2001, ApJ, 561, L93 Grindlay, J.E., Cool, A.M., Callanan, P.J., Baiylyn, C.D., Cohn, H.N., & Lugger, P.M. 1995, ApJ, 455, L47 Grindlay, J. E., Heinke, C.O., Edmonds, P. D., Murray, S. S., & Cool, A. M. 2001, ApJ, 563, L53 Harris, W. E. 1996, AJ, 112, 1487 Henry, W.G., & Fekel, F.C. 2002, PASP, 114, 988 Kaluzny, J. 1997, A&AS, 122, 1 Kaluzny, J., Olech, A., Stanek, K.Z. 2001, AJ, 121, 1533 Kaluzny, J., Rucinski, S.M., & Thompson, I.B. 2002, astro-ph/0209345 Kwee, K. K., & van Woerden, H. 1956, Bull. Astron. Inst. Netherlands, 12, 327 Lauzeral, C., Ortolani, S., Auri?re, M., Melnick, J. 1992, A&A, 262, 63 e Mochnacki, S.W., & Doughty, N.A., 1972, MNRAS, 156, 51 Piotto, G. et al. 2002, A&A, 391, 945 Plewa, T. 1988, Acta Astron., 38, 415 Olech, A., Wozniak, P.R., Alard, C., Kaluzny, J., Thompson, I.B. 1999, MNRAS, 310, 759 Ritter, H., & Kolb, U. 1998, A&AS, 129, 83 Reid, I. N., & Gizis, J. E. 1998, AJ, 116, 2929 Rodriguez, E., L?pez-Gonz?lez, M.J. 2000 A&A, 359, 597 o a Rucinski, S.M. 2000, AJ, 120, 319 Schwarzenberg-Czerny, A., 1989, MNRAS, 241, 153 Schwarzenberg-Czerny, A., 1996 ApJ, 460, L107 Stetson, P. B. 1987, PASP, 99, 191 Taylor, J. M., Grindlay, J. E., Edmonds, P. D., & Cool, A. M. 2001, ApJ, 553, L169

– 13 – Wilson, R. E. 1979, ApJ, 234, 1054 Zerbi, F.M., 2000, in “ Delta Scuti and Related Stars”, ASP Conf. Ser. 210, eds. M. Breger & M.H. Montgomery, p. 322

A This preprint was prepared with the AAS L TEX macros v5.0.

– 14 –

Fig. 1.— A ?nding chart for variable V24. North is up and east is to the left. The ?eld of view is 30 ′′ on a side. The variable is the W-S component of a blend.

– 15 –

Fig. 2.— Color-magnitude diagrams of NGC 6397 with the positions of the variables marked. SX Phe stars are labeled with triangles. Eclipsing binaries and CVs are labeled on the left ?gure. For SX Phe stars and CVs the positions correspond to average colors and magnitudes.

– 16 –

P1 300 AoV 200



0 0 1 2 3 4 5 frequency [1/d]

Fig. 3.— The power spectrum for the V -band light curve of the variable V12=CV1.

– 17 –

Fig. 4.— Phased V -band light curves of variable V12=CV1 for nights of UT May 1, 2, 3, 5, 6, and 7, and UT June 3, 2002 (from top to bottom).

– 18 –

Fig. 5.— Phased BV I light curves of variable V12=CV1. Note the same scale for all ?lters.

– 19 –

Fig. 6.— Time domain V -band light curve of V13=CV6 for the night of UT May 1, 2002

– 20 –

Fig. 7.— Phased BV I light curves of variable V13=CV6

– 21 –

Fig. 8.— Phased V -band light curves of variables V7 (upper) and V19.

– 22 –

Fig. 9.— Phased V -band light curve of variable V14

– 23 –

Fig. 10.— Phased BV I light curves of variable V18. Note that data for B and I ?lters are shifted.

– 24 –

Fig. 11.— The V -band light curve of V20 phased with a period of 0.860 days (bottom) and with a period of 0.436 days.

– 25 –

Fig. 12.— Phased V light curves of V17 (upper) and V24.

– 26 –

Fig. 13.— Light curve of V22 for the nights of UT May 7 and UT June 2, 2002

– 27 –

Table 1: VARIABLE STARS IN THE CENTRAL REGION OF NGC 6397 Name (1) V7 V10 V11 V12 V13 V14 V15 V16 V17 V18 V19 V20 V21 V22 V23 V24 R.A. (2) 40 43.74 40 37.43 40 43.95 40 41.42 40 48.82 40 46.31 40 45.24 40 44.44 40 43.63 40 43.45 40 44.66 40 41.51 40 41.40 40 41.02 40 39.21 40 38.97 Decl. (3) -53 40 35.6 -53 40 36.4 -53 40 40.9 -53 40 19.6 -53 39 49.0 -53 41 15.9 -53 40 25.2 -53 40 42.0 -53 41 16.8 -53 40 28.1 -53 40 23.8 -53 40 33.7 -53 40 23.9 -53 40 42.2 -53 40 46.9 -53 40 23.3 HST (4) W4(100,222) ID (5) WF4-2 BS11 BS9 U23, CV1 U10, CV6

17 17 17 17 17 17 17 17 17 17 17 17 17 17 17 17

W4(151,246) W1(489,504) W3(511,101) W4(548,348) W4(122,53) BS10 W4(190,223) MSP, WF4-1 W4(384,522) PC(345,91) PC(386,266) PC-1 PC(697,276) BS6 PC(559,441) BS7 BS8 BS16

Note: Cols. (2)-(3): Units of right ascension are hours, minutes and seconds, and units of declination are degrees, arcminutes, and arcseconds. Col. (5) Pixel coordinates on the HST archive image u5dr0401r proceeded by a name of WFPC-2 camera CCD. Col. (6) Other names of variables used in Grindlay et al. (2001), Taylor et al. (2001), and Lauzeral et al. (1992)

– 28 –

Table 2: PHOTOMETRIC DATA FOR NGC 6397 VARIABLES Name Type Period B V I ?V V7 W UMa 0.2699(2) 17.72 17.05 16.11 0.47 V10 SX Phe 0.03006 16.36 15.97 15.46 0.12 V11 SX Phe 0.03826 15.78 15.40 14.88 0.05 V12 CV 0.472(2) 18.51 17.95 16.96 0.37 V13 CV 0.2352(5) 20.04 19.35 18.27 0.45 V14 Eclipsing 0.3348(7) 20.21 19.19 17.93 1.02 V15 SX Phe 0.02145 15.80 15.44 14.94 0.05 V16 Ell 1.35406 17.36 16.65 15.71 0.15 V17 γ Dor? 0.525(5) 16.81 16.17 15.32 0.025 V18 Eclipsing 0.7871(3) 16.23 15.71 15.01 0.14 V19 W UMA 0.2538(2) 17.75 17.09 16.16 0.06 V20 W UMa?, γ Dor? 0.861(3) 16.22 15.75 15.08 0.04 V21 SX Phe 0.03896 15.88 15.48 14.91 0.30 V22 ? ? 16.61 16.16 15.58 0.11 V23 SX Phe 0.03717 16.05 15.66 15.14 0.04 V24 γ Dor? 0.457(2) 17.07 16.45 15.60 0.02 Note: Periods are given in days. BV I magnitudes are given at maximum brightness with exception of SX Phe stars for which average magnitudes are listed. The last column gives the di?erence between observed extremes for the V light curves: ?V = Vmin ? Vmax . For CVs the quoted magnitudes refer to 8 nights from the beginning of UT May, 2002.



All rights reserved Powered by 甜梦文库 9512.net

copyright ©right 2010-2021。