ArticlePDF Available

A large and homogeneous sample of SX Phe stars in the Fornax dwarf spheroidal galaxy .

Authors:
Mem. S.A.It. Vol. 77, 219
c
SAIt 2006 Memorie della
A large and homogeneous sample of SX Phe
stars in the Fornax dwarf spheroidal galaxy
E. Poretti1, L. Dell’Arciprete1, G. Clementini2, E. V. Held3, C. Greco2,4,
M. Gullieuszik3,5, M. Maio4, L. Rizzi6, M. Catelan7, H.A. Smith8, B. Pritzl9,
N. De Lee8, and A. Rest10
1INAF Osservatorio Astronomico di Brera, Via Bianchi 46, I-23807 Merate, Italy
2INAF Osservatorio Astronomico di Bologna, Via Ranzani 1, I-40127 Bologna, Italy
3INAF Osservatorio Astronomico di Padova, Vicolo dell’Osservatorio 5, I-35122
Padova, Italy
4Dipartimento di Astronomia, Universit`
a di Bologna, Via Ranzani 1, I-40127 Bologna,
Italy
5Dipartimento di Astronomia, Universit`
a di Padova, Vicolo dell’Osservatorio 2, I-35122
Padova, Italy
6Institute for Astronomy, University of Hawaii, 2680 Woodlawn Drive, Honolulu, HI,
USA
7Departamento de Astronom´
ıa y Astrof´
ısica, Pontificia Universidad Cat´
olica de Chile,
Avenida Vicuna Mackenna 4860, 782-0436 Macul, Santiago, Chile
8Department of Physics and Astronomy, Michigan State University, East Lansing, MI
48824-2320, USA
9Macalester College, 1600 Grand Avenue, Saint Paul, MN 55105, USA
10 Cerro Tololo Inter-American Observatory, Casilla 603, La Serena, Chile
Abstract. We report on the detection of sixty–one SX Phe stars in the Fornax dwarf
spheroidal galaxy using the Wide–Field Imager at the 2.2m ESO–MPI telescope. In spite of
their extreme faintness (22.0<B<24.0), the light curves of these short–period pulsators
have been clearly detected. We compare the Period–Luminosity relations in the Fornax
galaxy and in the Milky Way. The light curves of a candidate SX Phe star in the For 5
globular cluster is also presented (Magellan/Clay observations).
Key words. Stars: Population II Stars: variables : δSct Galaxies : Fornax Stars:
oscillations Galaxies: distances
1. Introduction on the nomenclature
of short–period pulsating stars
The nomenclature of the short–period
(P<0.20 d) pulsating stars located below the
Send oprint requests to: E. Poretti
Horizontal Branch in the classical instability
strip is a bit confusing. In the Milky Way there
is a physical distinction between δSct and
SX Phe stars: the former are Pop. I stars, the
latter are Pop. II stars. Among Pop. I stars,
the amplitude can vary from a few 0.001 mag
220 Poretti et al.: SX Phe stars in Fornax
to several 0.1 mag. Even if high amplitude δ
Sct (HADS) stars are likely radial pulsators,
there are several examples of the excitation of
nonradial modes (Poretti 2003). In a similar
way, SX Phe stars show a mixture of modes,
especially in globular clusters. Therefore, the
separation between HADS and SX Phe stars is
not an easy task in the Milky Way, if no details
about the metallicity are known. The study of
double–mode pulsators has introduced a new
tool, since the ratio between the fundamental
and the first overtone radial modes is sensitive
to metallicity eects (Poretti et al. 2005).
In the past, it was quite common to use
the term “Dwarf Cepheids” when these short–
period variables were observed in globular
clusters. This originates from the fact that they
can be used as distance indicators since they
follow a PLrelation similar to the more
evolved Classical Cepheids. However, the term
has an unclear physical meaning in the astro-
physical context, since the two groups of stars
are quite dierent. It is more appropriate to use
the SX Phe term to identify the short–period
variables in the globular clusters, since they are
most likely old, metal–poor stars. On the other
hand, it sounds inappropriate for the variables
in the galaxies, since they could have dierent
ages and metallicities; hence, we should ap-
ply the same distinction as in the Milky Way,
i.e., SX Phe and HADS stars. Such a distinc-
tion may seem a bit pedantic, so the old term
“Dwarf Cepheids” still survives in the current
literature.
The use of pulsating stars with period
shorter than 0.20 d as distance indicators has
made some progress in the last years, partic-
ularly in the case of globular clusters (e.g.,
Pych et al 2001, Mazur et al. 2003). On the
other hand, due to their faintness, not many
of these pulsating stars have been discovered
so far in other galaxies: Mateo et al. (1998)
and Poretti (1999) describe the first results ob-
tained on the Carina dwarf spheroidal (dSph)
galaxy. However, it appeared possible to ex-
ploit in a more complete way their potential-
ities by performing more intensive and deep
surveys. Therefore, we were strongly moti-
vated to start an observational project on the
Fornax dSph galaxy with the twofold goal to
reduce the gap between knowledge of SX Phe
stars in the Milky Way and those in other
galaxies and to test them as distance indicators.
2. Observations and data reduction
The feasibility of our project to detect SX Phe
stars in external galaxies was demonstrated by
the first observations of the northern part of the
Fornax dSph galaxy in November 2001. The
Wide–Field Imager (WFI) at the 2.2 m ESO–
MPI telescope was kept fixed on the same po-
sition on three consecutive nights. Since the
expected periods were very short, we took
three consecutive images in B–light (700 sec
each), followed by a single exposure in V–light
(1000 sec). This strategy ensured the dense
Btimeseries necessary to perform a reliable
frequency analysis and allowed us to obtain
the mean brightness and amplitude values in
a two–colour system. Photometric reductions
were carried out with the packages DAOPHOT
ALLSTAR-II (Stetson 1996) and ALLFRAME
(Stetson 1994). The variable stars were identi-
fied using the Image Subtraction Technique as
performed within the package ISIS 2.1 (Alard
2000). We have fully reduced four chips of the
WFI mosaic, detecting 61 short–period stars
and hundreds of RR Lyr variables. For the lat-
ter stars, see Clementini et al. (these proceed-
ings) and Greco et al. (2005).
The Fornax galaxy is known to host “a mix
of old and intermediate–age stars with dierent
metal abundances” (Held 2005) and therefore
we should distinguish between HADS and SX
Phe stars. However, at this stage, we consider
them all as SX Phe stars, i.e., stars belonging
to the older population. This seems justified by
the very short periods we found and by the av-
erage low metallicity of the Fornax galaxy. We
also stress that such a sample is very large and,
being comprised of similar stars at the same
distance, quite unique for number and homo-
geneity.
3. The detection of SX Phe stars
The intranight light curves are accurate enough
to exhibit the rapid variability of the SX Phe
Poretti et al.: SX Phe stars in Fornax 221
Fig. 1. The intranight light curves of three SX Phe
stars in the Fornax galaxy. Variability is discernible
on these curves even at the shortest periods (upper
row). Note the faintness of the variables.
Fig. 2. Folded light curves of SX Phe stars in the
Fornax galaxy. The Bmean luminosity increases by
1.7 mag from the shortest to longest periods.
stars (Fig. 1), in spite of their extreme faint-
ness (22.0<B<24.0). The timeseries just
sketch the light curves for the shortest peri-
ods (upper row, P=92 min; each exposure cov-
ers 0.13 p). They define the light curve shape
in a clearer way when the period lengthens
(middle row, P=114 min; the exposure cov-
ers 0.10 p) and, finally, they show the regular
light curve for the longest periods (lower row,
P=180 min; 0.06 p). When performing the fre-
Fig. 3. Folded light curves and power spectrum of
the candidate SX Phe star in the For 5 globular clus-
ter (Magellan/Clay observations).
quency analysis, the Bdata covering three con-
secutive nights ensured a high level of detec-
tion. Figure 2 shows the folded light curves in
B–light (61 datapoints); the Vdata (17 points)
have also been folded using the period val-
ues obtained from the more numerous Bdata.
Figure 2 is a PLrelation by itself: we can
see how the mean Bmagnitude brightens from
24.0 to 22.0 with increasing period.
In a follow–up of the project, we obtained
specific observations of the globular clusters in
Fornax using the Magellan telescope and the
Clay camera. Figure 3 shows the Band Vlight
curves (the Vfilter has been used to obtain the
denser timeseries here) and the power spec-
222 Poretti et al.: SX Phe stars in Fornax
Fig. 4. The solid line (least–squares fit) is the PL
relation as derived from Fornax SX Phe variables
(filled circles); the dashed line is the PLrelation as
obtained from SX Phe and HADS stars in the Milky
Way assuming (mM)V=20.72 for Fornax.
trum of the only candidate we found in For 5.
The period is 0.0495 d (i.e., 71 min) and the
full amplitude is 0.21 mag in V–light. We also
note that two SX Phe stars were found in the
outskirts of For 3 on the WFI images.
4. The PLrelation and the distance
modulus
We used the periods and the mVapparent mag-
nitudes to determine the slope of the Period–
Luminosity (PL) relation in Fornax (Fig. 4,
solid line). The result, 3.02 ±0.44, is smaller
than that observed in the Milky Way (–3.725,
McNamara 1997). We stress that the range
covered by the Fornax stars in log P(from
1.35 to 0.90) is narrower than that covered
by the Milky Way stars (from 1.40 to 0.55).
The dierent distribution of periods probably
reflects the dierences in the stellar content:
the Milky Way contains many more Pop. I
HADS stars than Fornax and in turn these stars
have longer periods than Pop. II SX Phe stars.
The PLrelation given by McNamara
(1997) can also be drawn in the log PmV
plane by assuming a distance modulus of (m
M)V=20.72 for Fornax (dashed line), as
derived from RR Lyr variables (Greco et al.
2005). As can be seen (Fig. 4), the two lines
are nearly coincident in the period range of the
Fornax variables. We can also make the reverse
exercise to calculate the distance modulus of
Fornax assuming the McNamara PLrelation.
In such a case we get (mM)V=20.71 ±0.04,
a value in excellent agreement with the previ-
ous ones (e.g., Buonanno et al. 1999, Saviane
et al. 2000, Greco et al. 2005).
5. Conclusions
The large sample of SX Phe variables in the
Fornax galaxy shows a well–defined PL
relation, which is in good agreement with
the one found in the Milky Way. The results
briefly presented here seem to confirm that
these short–period stars can be used as stellar
candles in the Local Group and that they can
provide an independent method to measure dis-
tances.
References
Alard, C. 2000, A&AS, 144, 363
Buonanno, R. et al., 1999, AJ, 118, 1671
Greco C. et al. 2005, in “Resolved Stellar
Populations”, D. Valls-Gabaud and
M. Chavez Eds., ASP Conf. Series, in
press [astro-ph/0507244]
Held, E. 2005, in “Near Field Cosmology with
Dwarf Elliptical Galaxies”, IAU Colloq.
198, B. Binggeli and H. Jerjen Eds, in press
[astro-ph/0506197]
Mateo, M., Hurley-Keller, D., & Nemec, J.
1998, AJ, 115, 1856
Mazur, B., Krzemi´
nski, W. Thompson, I.B.
2003, MNRAS, 340, 1205
McNamara, D.H. 1997, PASP, 109, 1221
Poretti, E. 1999, A&A, 345, 385
Poretti, E. 2003, A&A, 409, 1031
Poretti, E. et al. 2005, A&A, 440, 1097
Pych, W. et al. 2001, A&A, 367, 148
Saviane, I., Held, E.V. Bertelli, G. 2000,
A&AS, 355, 56
Stetson, P.B. 1994, PASP, 106, 250
Stetson, P.B. 1996, User’s Manual for
DAOPHOT II
Article
Full-text available
Variable stars have been identified for the first time in Fornax 4, the globular cluster located near the center of the Fornax dwarf spheroidal galaxy. By applying the image-subtraction technique to B and V time series photometry obtained with the MagIC camera of the 6.5 m Magellan Clay telescope and with the wide field imager of the 4 m Blanco CTIO telescope, we detected 27 RR Lyrae stars (22 fundamental-mode, 3 first-overtone, and 2 double-mode pulsators) in a 2.4' × 2.4' area centered on Fornax 4. The average and minimum periods of the ab-type RR Lyrae stars, Pab = 0.594 days and Pab,min = 0.5191 days, respectively, as well as the revised position of the cluster in the horizontal branch type-metallicity plane, all consistently point to an Oosterhoff-intermediate status for the cluster, unlike what is seen for the vast majority of Galactic globular clusters, but in agreement with previous indications for the other globular clusters in Fornax. The average apparent magnitude of the RR Lyrae stars located within 30'' from the cluster center is V(RR) = 21.43 ± 0.03 mag (σ = 0.10 mag, average of 12 stars), leading to a true distance modulus of μ0 = 20.64 ± 0.09 mag or μ0 = 20.53 ± 0.09 mag, depending on whether a low ([Fe/H] = -2.0) or a moderately high ([Fe/H] = -1.5) metallicity is adopted.
Article
Full-text available
We have carried out a photometric and spectroscopic survey of bright high-amplitude δ Scuti (HADS) stars. The aim was to detect binarity and multiperiodicity (or both) in order to explore the possibility of combining binary star astrophysics with stellar oscillations. Here, we present the first results for 10, predominantly southern, HADS variables. We detected the orbital motion of RS Gru with a semi-amplitude of ∼6.5 km s−1 and 11.5 d period. The companion is inferred to be a low-mass dwarf star in a close orbit around RS Gru. We found multiperiodicity in RY Lep from both photometric and radial velocity data and detected orbital motion in the radial velocities with hints of a possible period of 500–700 d. The data also revealed that the amplitude of the secondary frequency is variable on the time-scale of a few years, whereas the dominant mode is stable. Radial velocities of AD CMi revealed cycle-to-cycle variations, which might be due to non-radial pulsations. We confirmed the multiperiodic nature of BQ Ind, while we obtained the first radial velocity curves of ZZ Mic and BE Lyn. The radial velocity curve and the O–C diagram of CY Aqr are consistent with the long-period binary hypothesis. We took new time series photometry on XX Cyg, DY Her and DY Peg, with which we updated their O–C diagrams.
Article
Full-text available
Preliminary evidence is presented reaffirming that SX Phe, RR Lyrae, and Type II Cepheid variables may be characterized by a common Wesenheit period-magnitude relation, to first order. Reliable distance estimates to RR Lyrae variables and Type II Cepheids are ascertained from a single VI-based reddening-free relation derived recently from OGLE photometry of LMC Type II Cepheids. Distances are computed to RR Lyrae (d~260 pc), and variables of its class in the galaxies IC 1613, M33, Fornax dSph, LMC, SMC, and the globular clusters M3, M15, M54, omega Cen, NGC 6441, and M92. The results are consistent with literature estimates, and in the particular cases of the SMC, M33, and IC 1613, the distances agree with that inferred from classical Cepheids to within the uncertainties: no corrections were applied to account for differences in metallicity. Moreover, no significant correlation was observed between the distances computed to RR Lyrae variables in omega Cen and their metallicity, despite a considerable spread in abundance across the sample. In sum, concerns regarding a sizeable metallicity effect are allayed when employing VI-based reddening-free Cepheid and RR Lyrae relations. Comment: Accepted for Publication (JAAVSO)
Article
Full-text available
New candidate variable stars have been identified in the Small Magellanic Cloud cluster NGC121, by applying both the image subtraction technique (ISIS, Alard 2000) and the Welch & Stetson (1993) detection method to HST WFPC2 archive and ACS proprietary images of the cluster. The new candidate variable stars are located from the cluster's Main Sequence up to Red Giant Branch. Twenty-seven of them fall on the cluster Horizontal Branch and are very likely RR Lyrae stars. They include the few RR Lyrae stars already discussed by Walker & Mack (1988). We also detected 20 Dwarf Cepheid candidates in the central region of NGC121. Our results confirm the "true" globular cluster nature of NGC121, a cluster that is at the young end of the Galactic globulars' age range. Comment: 3 pages, 3 figures
Article
Full-text available
We have discovered 20 dwarf Cepheids (DC) in the Carina dSph galaxy from the analysis of individual CCD images obtained for a deep photometric study of the system. These short-period pulsating variable stars are by far the most distant (~100 kpc) and faintest (V ~ 23.0) DCs known. The Carina DCs obey a well-defined period-luminosity relation, allowing us to readily distinguish between overtone and fundamental pulsators in nearly every case. Unlike RR Lyr stars, the pulsation mode turns out to be uncorrelated with light-curve shape, nor do the overtone pulsators tend towards shorter periods compared to the fundamental pulsators. Using the period-luminosity (PL) relations from Nemec et al. (1994 AJ, 108, 222) and McNamara (1995, AJ, 109, 1751), we derive (m-M)_0 = 20.06 +/- 0.12, for E(B-V) = 0.025 and [Fe/H] = -2.0, in good agreement with recent, independent estimates of the distance/reddening of Carina. The error reflects the uncertainties in the DC distance scale, and in the metallicity and reddening of Carina. The frequency of DCs among upper main sequence stars in Carina is approximately 3%. The ratio of dwarf Cepheids to RR Lyr stars in Carina is 0.13 +/- 0.10, though this result is highly sensitive to the star-formation history of Carina and the evolution of the Horizontal Branch. We discuss how DCs may be useful to search effectively for substructure in the Galactic halo out to Galactocentric distances of ~100 kpc. Comment: 20 pages of text, 7 figures
Article
Strömgren _uvbyβ photometry of SX Phoenicis and large-amplitude δ Scuti variables is analyzed to determine the mean temperatures, metal abundances, and surface gravities of the stars. The mean temperatures and fundamental periods are used in conjunction with the pulsation equation and stellar models to derive masses, radii, surface gravities, and M_bol of the variables. The stars exhibit a period-luminosity relation. With the aid of Hipparcos trigonometric parallaxes the zero-point of the M_v, p is set. We find M_v = -3.725 log P - 1.933. This P-L relation is utilized to find the M-v values of RR Lyrae stars or horizontal-branch stars in globular clusters and the Carina galaxy-objects that contain both SX Phe, and RR Lyrae variables, or horizontal-branch stars. The M_v values of the metal-poor RR Lyrae stars are found to be consistent with M_v [Fe/H] calibrations derived by Sandage (1993) and by McNamara (1997) from revised Baade- Wesselink M_v values. At [Fe/H] = -1.9 M_v is 0.42. Feast and Catchpole (1997) suggestion that M_v = 0.25 at [Fe/H] = -1.9 for RR Lyrae is discussed. This M_v value is too luminous because it makes the (T_eff) of RR Lyraes too high-well off any (T_eff) = _f (color index) calibration. (SECTION: Stars)
Article
The central brightness cusps seen in some globular clusters are thought to be the relics of a gravothermal core collapse that occurred sometime in the clusters' past. Recent observations show that the centers of such clusters are bluer than their outskirts, indicating that the stellar populations there are somehow different than those farther out, presumably as a result of unusual physical processes that took place in these extremely dense regions. Here I analyze a large body of digital imagery from the Canada-France-Hawaii Telescope and the Hubble Space Telescope to obtain color-magnitude and color-color diagrams for stars in the central two arcminutes of the prototypical core-cusp globular cluster M15 = NGC 7078 = C 2127 + 119. These data were reduced with a new computer program, named ALLFRAME, that is described in detail here for the first time. ALLFRAME makes simultaneous use of the geometric and photometric information from all images of a given field to derive a self-consistent set of positions and magnitudes for all detected starlike objects in that area of sky, thereby extending the range of magnitude and crowding conditions for which useful photometry is obtainable. I tentatively conclude that the color gradient in M15 is due to three distinct effects: (1) there is a deficiency of the brightest red giants in the central regions of the cluster; (2) the giant branch of the dominant cluster population shifts systematically toward the blue as the center of the cluster is approached; and (3) the very center of the cluster (radius approximately equal or less than 12 seconds) contains a large population of blue and yellow stragglers that occupy the area of the color-magnitude diagram between the main-sequence turnoff and the horizontal branch and between the extended blue horizontal branch and the subgiant branch; many of these appear to have a significant ultraviolet excess.
Conference Paper
We present the deepest optical CMD to date of M 32, obtained from HST ACS/HRC images. The dominant feature is the RC, whose CMD location suggests a mean age between 8 and 10 Gyr for [Fe/H] = -0.2 in M 32. We detect for the first time the RGB and AGB bumps in M 32 which constrain its mean age to be 5-10 Gyr old at 2' from the center. Bright AGB stars indicate the presence of intermediate-age populations. A detected blue component of stars may indicate for the first time the presence of a young stellar population of ~0.5 Gyr in our M 32 field. However, it is likely that the younger stars of this blue plume belong to M 31 rather than to M 32. The fainter stars populating the blue plume indicate the presence of stars not younger than 1 Gyr and/or blue straggler stars in M 32.
Article
We report on the discovery of 11 multimode SX Phoenicis type blue stragglers in the field of the southern globular cluster NGC 3201. In these variables both radial and non-radial modes are excited. For three variables the derived period ratio is close to that observed in SX Phoenicis itself, suggesting that these stars are pulsating in the fundamental and the first-overtone radial modes. Using the McNamara period—luminosity relation we have estimated the apparent distance modulus to NGC 3201 to be 14.08 ± 0.06 ± 0.1 mag.
  • D H Mcnamara
McNamara, D.H. 1997, PASP, 109, 1221
  • E Poretti
Poretti, E. et al. 2005, A&A, 440, 1097
Near Field Cosmology with Dwarf Elliptical Galaxies
  • E Held
  • Colloq
Held, E. 2005, in " Near Field Cosmology with Dwarf Elliptical Galaxies ", IAU Colloq. 198, B. Binggeli and H. Jerjen Eds, in press [astro-ph/0506197]
User's Manual for DAOPHOT II
  • P B Stetson
Stetson, P.B. 1996, User's Manual for DAOPHOT II