The OGLE Collection of Variable Stars. Over 45 000 RR Lyrae Stars in the Magellanic System
I. Soszyński, A. Udalski, M. K. Szymański, Ł. Wyrzykowski, K. Ulaczyk, R. Poleski, P. Pietrukowicz, S. Kozłowski, D. Skowron, J. Skowron, P. Mróz, M. Pawlak
aa r X i v : . [ a s t r o - ph . S R ] J u l ACTA ASTRONOMICA
Vol. (2016) pp. 131–147 The OGLE Collection of Variable Stars.Over 45 000 RR Lyrae Stars in the Magellanic System ∗ I. S o s z y ´n s k i , A. U d a l s k i , M. K. S z y m a ´n s k i , Ł. W y r z y k o w s k i ,K. U l a c z y k , , R. P o l e s k i , , P. P i e t r u k o w i c z , S. K o z ł o w s k i ,D. M. S k o w r o n , J. S k o w r o n , P. M r ó z and M. P a w l a k Warsaw University Observatory, Al. Ujazdowskie 4, 00-478 Warszawa, Polande-mail: (soszynsk,udalski)@astrouw.edu.pl Department of Physics, University of Warwick, Gibbet Hill Road, Coventry,CV4 7AL, UK Department of Astronomy, Ohio State University, 140 W. 18th Ave., Columbus,OH 43210, USA
Received June 21, 2016
ABSTRACTWe present the largest collection of RR Lyrae stars in the Magellanic System and in its fore-ground. The sample consists of 45 451 RR Lyr stars, of which 39 082 were detected toward the LargeMagellanic Cloud and 6369 toward the Small Magellanic Cloud. We provide long-term time-seriesphotometric measurements collected during the fourth phase of the Optical Gravitational LensingExperiment (OGLE-IV).We discuss several potential astrophysical applications of our collection: investigation of thestructure of the Magellanic Clouds and the Galactic halo, studies of the globular clusters in the Mag-ellanic System, analysis of double-mode RR Lyr stars, and search for RR Lyr stars in eclipsing binarysystems.
Key words:
Stars: variables: RR Lyrae – Stars: oscillations – Stars: Population II – MagellanicClouds – Catalogs
1. Introduction
RR Lyrae stars are a powerful tool to trace the oldest ( ≥
10 Gyr) stellar compo-nent in our and other galaxies. These radially pulsating stars are numerous, easilyidentifiable, and present in all Local Group galaxies, irrespective of their morpho-logical type. RR Lyr stars are standard candles making them important distance ∗ Based on observations obtained with the 1.3-m Warsaw telescope at the Las Campanas Observa-tory of the Carnegie Institution for Science. A. A. indicators, probes of the three-dimensional structures of their parent galaxies, andtracers of the star formation history. Large samples of RR Lyr variables have provenuseful in investigating the metallicity distribution in galaxies and in determinationof the interstellar extinction maps.The Optical Gravitational Lensing Experiment (OGLE, Udalski, Szyma´nskiand Szyma´nski 2015) has already released catalogs of RR Lyr stars in the Large(LMC) and Small Magellanic Clouds (SMC) based on the observations obtained in1997–2000 during the OGLE-II (Soszy´nski et al. et al. et al. etal. et al. et al. et al. et al. et al. et al. et al. et al. et al. et al. et al. et al. et al. ol. 66
2. Observations and Data Reduction
The time-series photometry used in this study has been obtained with the 1.3-mWarsaw telescope at Las Campanas Observatory (operated by the Carnegie Insti-tution for Science), Chile, between March 2010 and July 2015. The telescope isequipped with a mosaic camera composed of 32 CCDs, each with 2048 by 4096pixels, providing a field of view of 1.4 square degrees on the sky. Most of the ob-servations were obtained through the Cousins I filter – typically from 100 to 750points, depending on the field. In the Johnson V -band we secured from several to260 observations for color information.Altogether 475 OGLE-IV fields cover about 650 square degrees in the Mag-ellanic System, including the Magellanic Bridge between the LMC and SMC andselected peripheral areas, up to 20 degrees from the centers of the galaxies. The to-tal number of point sources in the Magellanic Cloud OGLE-IV database exceeds 75million. The OGLE data reduction pipeline is based on the Difference Image Anal-ysis technique (Alard and Lupton 1998, Wo´zniak 2000). The reduction procedures,photometric calibrations and astrometric transformations have been described byUdalski et al. (2015a).
3. Selection and Classification of RR Lyrae Stars
We performed a period search for nearly all I -band light curves stored in theOGLE database. The only cut was done on the number of data points that had to belarger than 30. We used the F NPEAKS code † which calculates the Fourier amplitudespectra and provides the best periods with their signal-to-noise ratios. For each starwe derived two periods, the second one on the residual light curve obtained by thesubtraction of the primary periodicity.A search for RR Lyr stars was conducted on light curves with periods between0.2 and 1 day. The preselection of the candidates was based on the Fourier decom-position of the light curves (Simon and Lee 1981) and the template fitting to the I -band light curves. Double-mode RR Lyr stars were identified on the basis of theirperiod ratios, which fall in a narrow range of 0.72–0.75. However, the automaticalgorithms played only a supporting role in the process of the manual selection andclassification of RR Lyr stars. The final decision on each object was made aftera visual inspection of its light curve. In doubtful cases we took into account theposition of the star in the color–magnitude diagram, as well as period–luminosity,period–amplitude, and other diagrams.The selected sample of RR Lyr variables has been divided into three classes:fundamental-mode RRab stars, first-overtone RRc stars, and double-mode RRdstars. Our classification was based on the periods, amplitudes, and light curveshapes of the stars. In several dozen cases we corrected the pulsation modes pro-vided by Soszy´nski et al. (2009, 2010). All objects classified in the OGLE-III † http://helas.astro.uni.wroc.pl/deliverables.php?lang=en&active=fnpeaks A. A. catalogs as RRe stars (second-overtone pulsators) have been incorporated to theRRc group. The second-overtone RR Lyr stars have not been separated from thefirst-overtone variables because of the doubts whether the RRe stars exist at all.We did not find any natural boundary, which could be used for an unambiguousseparation of the first- and second-overtone RR Lyr stars.As a result, we found 18 158 RR Lyr stars in both Clouds that were not recordedduring previous stages of the OGLE survey. Only 227 of them (less than 1% of theOGLE-III sample) were found in the region covered by the OGLE-III fields, con-firming the high completeness of the OGLE collection of variable stars. Othernewly detected variables lie in the outer regions of the two galaxies. We also ver-ified the OGLE-III samples of RR Lyr stars (Soszy´nski et al. ≈
4. RR Lyrae Stars in the Magellanic System
The current version of the OGLE collection of RR Lyr stars in the MagellanicSystem contains variables detected during the previous stages of the OGLE project(Soszy´nski et al. etal. h ftp://ftp.astrouw.edu.pl/ogle/ogle4/OCVS/lmc/rrlyr/ftp://ftp.astrouw.edu.pl/ogle/ogle4/OCVS/smc/rrlyr/http://ogle.astrouw.edu.pl ol. 66 T a b l e 1
Reclassified variables from the OGLE-III catalogs of RR Lyr stars in the Magellanic CloudsIdentifier New Identifier Newclassification classificationOGLE-LMC-RRLYR-00077 Other OGLE-LMC-RRLYR-13259 EclipsingOGLE-LMC-RRLYR-00485 Other OGLE-LMC-RRLYR-13512 EclipsingOGLE-LMC-RRLYR-00803 Spotted OGLE-LMC-RRLYR-15806 OtherOGLE-LMC-RRLYR-00824 Other OGLE-LMC-RRLYR-16124 EclipsingOGLE-LMC-RRLYR-00961 Eclipsing OGLE-LMC-RRLYR-16210 EclipsingOGLE-LMC-RRLYR-01104 Eclipsing OGLE-LMC-RRLYR-16426 OtherOGLE-LMC-RRLYR-01257 Other OGLE-LMC-RRLYR-16656 EclipsingOGLE-LMC-RRLYR-01802 Eclipsing OGLE-LMC-RRLYR-16795 EclipsingOGLE-LMC-RRLYR-02171 Other OGLE-LMC-RRLYR-17073 EclipsingOGLE-LMC-RRLYR-02376 Other OGLE-LMC-RRLYR-17117 EclipsingOGLE-LMC-RRLYR-02390 Eclipsing OGLE-LMC-RRLYR-17396 EclipsingOGLE-LMC-RRLYR-02548 Eclipsing OGLE-LMC-RRLYR-17584 EclipsingOGLE-LMC-RRLYR-03158 Eclipsing OGLE-LMC-RRLYR-17847 OtherOGLE-LMC-RRLYR-03802 Eclipsing OGLE-LMC-RRLYR-18086 OtherOGLE-LMC-RRLYR-04103 Eclipsing OGLE-LMC-RRLYR-18360 OtherOGLE-LMC-RRLYR-04426 Classical Cep. OGLE-LMC-RRLYR-18854 OtherOGLE-LMC-RRLYR-04733 Other OGLE-LMC-RRLYR-19067 EclipsingOGLE-LMC-RRLYR-04862 Eclipsing OGLE-LMC-RRLYR-19207 OtherOGLE-LMC-RRLYR-04892 Eclipsing OGLE-LMC-RRLYR-19243 EclipsingOGLE-LMC-RRLYR-05128 Eclipsing OGLE-LMC-RRLYR-19258 OtherOGLE-LMC-RRLYR-05282 Other OGLE-LMC-RRLYR-19438 EclipsingOGLE-LMC-RRLYR-05305 Eclipsing OGLE-LMC-RRLYR-20089 OtherOGLE-LMC-RRLYR-05784 Eclipsing OGLE-LMC-RRLYR-20767 EclipsingOGLE-LMC-RRLYR-06232 Eclipsing OGLE-LMC-RRLYR-20781 OtherOGLE-LMC-RRLYR-06645 Other OGLE-LMC-RRLYR-20821 EclipsingOGLE-LMC-RRLYR-07073 Other OGLE-LMC-RRLYR-21161 EclipsingOGLE-LMC-RRLYR-07569 Eclipsing OGLE-LMC-RRLYR-21207 OtherOGLE-LMC-RRLYR-07905 Eclipsing OGLE-LMC-RRLYR-21255 EclipsingOGLE-LMC-RRLYR-07935 Other OGLE-LMC-RRLYR-21285 EclipsingOGLE-LMC-RRLYR-08281 Other OGLE-LMC-RRLYR-21455 EclipsingOGLE-LMC-RRLYR-08457 Other OGLE-LMC-RRLYR-22035 OtherOGLE-LMC-RRLYR-09009 Eclipsing OGLE-LMC-RRLYR-22482 EclipsingOGLE-LMC-RRLYR-09044 Eclipsing OGLE-LMC-RRLYR-22492 OtherOGLE-LMC-RRLYR-09614 Other OGLE-LMC-RRLYR-23055 OtherOGLE-LMC-RRLYR-10747 Other OGLE-LMC-RRLYR-23485 OtherOGLE-LMC-RRLYR-10933 Eclipsing OGLE-LMC-RRLYR-23513 EclipsingOGLE-LMC-RRLYR-10994 Eclipsing OGLE-LMC-RRLYR-23517 EclipsingOGLE-LMC-RRLYR-11143 Other OGLE-LMC-RRLYR-23685 OtherOGLE-LMC-RRLYR-11169 Eclipsing OGLE-LMC-RRLYR-23868 OtherOGLE-LMC-RRLYR-11606 Other OGLE-LMC-RRLYR-24247 OtherOGLE-LMC-RRLYR-12072 Eclipsing OGLE-LMC-RRLYR-24338 OtherOGLE-LMC-RRLYR-12151 Eclipsing OGLE-LMC-RRLYR-24428 EclipsingOGLE-LMC-RRLYR-12343 Other OGLE-SMC-RRLYR-0251 OtherOGLE-LMC-RRLYR-12875 Eclipsing OGLE-SMC-RRLYR-0720 Other A. A.
Each RR Lyr star has a unique identifier which follows the scheme introducedin the OGLE-III catalogs. The identifiers from OGLE-LMC-RRLYR-00001 toOGLE-LMC-RRLYR-24906 (in the LMC) and from OGLE-SMC-RRLYR-0001to OGLE-SMC-RRLYR-2475 (in the SMC) are reserved for the RR Lyr stars pre-sented by Soszy´nski et al. (2009, 2010). The identifiers with higher numbers areassigned to the newly detected RR Lyr stars in order of increasing right ascension.Our collection contains not only the most important parameters of the sources(their coordinates, modes of pulsation, periods, mean magnitudes in the I - and V -bands, amplitudes, and Fourier coefficients of the light curve decomposition), butalso the time-series VI photometry collected from the beginning of the OGLE-IVsurvey. This photometry can be combined with the OGLE-III and OGLE-II lightcurves (if available) from the Soszy´nski et al. (2009, 2010) catalogs, however inindividual cases one should compensate possible differences in the mean bright-ness and amplitudes between the previous and present stages of the OGLE project.About 8% of the RR Lyr stars included in the OGLE-III catalogs do not haveOGLE-IV photometry, mostly because they fell in technical gaps between CCDchips of the OGLE-IV mosaic camera. The parameters of these variables werecopied from the OGLE-III catalog. In the future, we plan to obtain the OGLE-IV time-series photometry also for the stars in the gaps, since these regions areobserved from time to time due to imperfections of the telescope pointing.
5. Completeness of the Sample
Due to a partial overlap of the adjacent OGLE-IV fields, some of the sourceswere recorded twice, independently in both fields. However, our collection containsonly one entry per star – in the case of the double detections we usually chose theone with the larger number of observing points in its light curve. These independentidentifications of the same RR Lyr stars may be used to estimate the completenessof our sample.We expect that the OGLE collection of RR Lyr variables is practically completein the central regions of the LMC and SMC, that were monitored since 2001 bythe OGLE-III and OGLE-IV surveys. The outer regions are affected by the gapsbetween CCD detectors of the mosaic camera, which reduce the completeness byabout 7%. The efficiency of our search for RR Lyr stars in the area covered by thepixels may be judged on the basis of the double detections. Outside the OGLE-IIIfields, 1284 variables from our sample had two entries in the OGLE-IV database(assuming that both light curves must have at least 100 points), so we had a chanceto find 2568 counterparts. We independently identified 2480 of them, which impliesthe completeness of about 96%.The highest completeness is expected for RRab stars, due to their characteristicsawtooth-like light curves. Indeed, the same method applied to the fundamental-mode pulsators gives the completeness well above 98%. For RRc and RRd stars ol. 66 et al. (2001). Our sample does not include 131 of theseobjects, of which 40 are not present in the OGLE-IV database (most of them lies inthe gaps between the CCD chips). The remaining 91 stars classified by MACHO asRR Lyr stars either clearly belong to other types of variable sources or are constantstars.Kim et al. (2014) published a list of periodic variable star candidates detectedfrom the EROS-2 LMC photometric database. These objects were classified usingan automatic random forest algorithm. The list of potential RR Lyr stars contains6607 sources not discovered during the previous stages of the OGLE survey or theMACHO project. We cross-matched an early version of our collection of RR Lyrstars in the LMC with the sample published by Kim et al. (2014) and we foundthat as many as 4408 object were missed in our list. For 3234 of these stars wefound their counterparts in the OGLE-IV database within 1 arcsec search radius.We carefully analyzed the light curves of these objects and found that 149 of themindeed are probable RR Lyr stars. Most of these overlooked variables turned outto be RRc stars with noisy, nearly sinusoidal light curves, sometimes affected by asmall number of points in their light curves. We supplemented our collection withthese newly identified RR Lyr variables. In turn, we do not confirm the Kim etal. (2014) classification for the remaining 3085 sources. For the majority of theseobjects we had no doubt that we deal with eclipsing binaries, d Sct stars, Cepheids,or simply just constant stars. There is also a number of sources in this group forwhich the OGLE light curves are too noisy to unambiguously categorize their typeof variability.
6. Discussion
The present version of the OGLE collection of RR Lyr stars in the MagellanicClouds is larger and purer than any other catalog of these pulsators detected in anyother environment. Therefore, our sample is an ideal tool to study RR Lyr starsthemselves, as well as the structure of the Magellanic Clouds and their interactionswith each other and our Galaxy. Below we present a few possible applications ofour collection, however we are far from being exhaustive.
RR Lyr stars are primary tracers of the ancient stellar population. These pulsat-ing stars are common in various environments, can be easy identified in time-seriessky surveys, and are standard candles, so can be used to study the distribution ofthe old population in three dimensions. The OGLE-III catalogs of RR Lyr stars(Soszy´nski et al.
A. A.
Fig. 1. Spatial distribution of RR Lyr stars in the OGLE fields toward the Magellanic Clouds.
Upperpanel presents members of the Magellanic Clouds (fainter group). Gray circles indicate positions oftwelve globular clusters that host RR Lyr stars.
Lower panel shows positions of the brighter groupof RR Lyr stars – consisting mostly of the Milky Way members. The boundary between fainter andbrighter groups has been adopted at 1 magnitude above the mean period–luminosity relations forRRab, RRc, and RRd stars, separately for the LMC and SMC. Additionally, blended stars have beenremoved from the bright group. Gray area shows the sky coverage of the OGLE fields. ol. 66 et al. et al. et al. W I = I − . ( V − I ) , separately for RRab, RRc, and RRd stars in the LMC andSMC. In the case of the brighter group, we also cleaned the sample from blendedvariables. In this procedure we relied on the light curve amplitudes, which aresmaller in blended variables than in typical pulsators with the same periods.The fainter RR Lyr stars belong mainly to the Magellanic Clouds, while thebrighter group is populated mostly by members of the Milky Way halo, althoughthere is no clear boundary between outer regions of these three galaxies. This isseen in the distribution of the brighter RR Lyr stars, which is roughly uniform overthe OGLE fields with the exception of the center of the LMC. The excess of brightRR Lyr stars in this region may be partially explained by the contamination of notremoved blends, but the majority of these RR Lyr stars have typical amplitudes oftheir light curves, so they do not seem to be substantially blended. Thus, these starsare likely located in the outskirts of the LMC stellar halo tidally stretched towardour Galaxy.The fainter group (upper panel of Fig. 1) mostly belong to the MagellanicClouds. The projection of the SMC halo on the celestial sphere seem to be round,while the LMC halo is obviously elongated. Moreover, the distribution of the LMCRR Lyr stars probably cannot be described by a simple ellipsoid, because the num-ber of RR Lyr stars in the North-East part of the LMC seems to be larger than in theopposite side. A detailed analysis of the three-dimensional distribution of RR Lyrstars in the Magellanic Clouds on the basis of our collection will be presented inthe forthcoming paper (Jacyszyn-Dobrzeniecka et al. in preparation). The Oosterhoff dichotomy observed in Galactic globular clusters is not presentamong the globular clusters in nearby dwarf galaxies, in particular in the LMC.Five of the LMC clusters (NGC 1466, NGC 1853, NGC 2019, NGC 2210, andNGC 2257) have properties that place them inside the zone of avoidance betweenthe two Oosterhoff groups in the Milky Way. This fact poses a significant challengeto the models assuming hierarchical merger formation of the Galactic halo.The catalog of extended objects in the Magellanic System by Bica et al. (2008)lists 18 globular clusters in the Magellanic Clouds, 14 of which are currently mon-40
A. A. itored by OGLE. We found RR Lyr stars in all but two of the observed globularclusters. We found no RR Lyr stars in Hodge 11 in the LMC and Lindsay 1 in theSMC (strictly speaking we identified one RR Lyr star in Hodge 11, but it is prob-ably a field variable located by chance in the area outlined by the cluster radius).The simplest explanation for the lack of RR Lyr stars in Hodge 11 and Lindsay 1 isthat these clusters are younger than ≈
10 Gyr.
T a b l e 2
Globular clusters containing RR Lyr starsCluster name RA Dec Cluster N RR N fieldRR (J2000) (J2000) radius [ ′ ] (estimated)NGC 121 00 h m s − ◦ ′ ′′ h m s − ◦ ′ ′′ h m s − ◦ ′ ′′ h m s − ◦ ′ ′′ h m s − ◦ ′ ′′ h m s − ◦ ′ ′′ h m s − ◦ ′ ′′ h m s − ◦ ′ ′′ h m s − ◦ ′ ′′ h m s − ◦ ′ ′′ h m s − ◦ ′ ′′ h m s − ◦ ′ ′′ Table 2 summarizes the properties of twelve globular clusters in the LMC andSMC that host RR Lyr stars. The coordinates and angular radii of the clusters aretaken from Bica et al. (2008). In the last two columns we provide numbers ofRR Lyr stars detected within one radius from the clusters’ centers and estimatednumbers of field RR Lyr stars that are expected to fall inside the same area. Weestimated the number of field variables counting RR Lyr stars in the rings from1.5 to 2.5 radii around the cluster centers and rescaling these numbers to the areaoccupied by clusters. The full lists of RR Lyr stars found within the cluster radiiare provided in the FTP site in the file gc.dat .In two clusters – NGC 1928 and NGC 1939 – we found only eight and sevenRR Lyr stars, respectively, while the expected number of field variables is about fivein both cases. Therefore, it cannot be excluded that all RR Lyr stars detected insidethe area outlined by the radii of these clusters are field variables. Spectroscopic andastrometric follow-up observations of these stars should give a definitive answer tothe question of their membership. In other globular clusters listed in Table 2 theidentification of a significant number of RR Lyr stars is firm, although in most casescluster members and field variables cannot be unambiguously distinguished. ol. 66
Fig. 2. Period distributions of RR Lyr stars in the Magellanic Clouds’ globular clusters. Blue, red,and green contours show histograms for RRab, RRc, and RRd (first-overtone periods) stars. Globularclusters are arranged by increasing metallicities, [Fe/H], given in the top right corner of each panel . The distribution of pulsation periods of RR Lyr stars in ten the richest clustersare presented in Fig. 2. The clusters are arranged by increasing metallicities toshow the progression of the period distribution with metallicity. RRd stars havebeen detected in four globular clusters and it is interesting that all of them haveintermediate metal abundances − . ≤ [ Fe / H ] ≤ − .
78 and fall in the Oosterhoffgap.42
A. A.
RR Lyr stars with two first radial modes simultaneously excited (RRd stars)constitute 5% of the total sample in the LMC and 10% in the SMC. These are thelargest sets of RRd stars known in any stellar environment, so they may serve asimportant testbeds for theories of stellar pulsation. Petersen diagram (period ratiosplotted against the longer period) is a sensitive tool widely used in asteroseismology(Popielski et al.
LMC SMC
Fig. 3. Petersen diagrams for double-mode RR Lyr variables in the LMC ( left panel ) and SMC ( rightpanel ). Black dots represent “classical” RRd stars, while empty circles mark anomalous RRd stars(Soszy´nski et al.
Fig. 3 shows the Petersen diagrams for RRd stars in the LMC and SMC. Thevast majority of double-mode RR Lyr stars has period ratios within a narrow rangeof 0 . < P / P F < .
75 and forms a curved sequence in the diagram. The se-quence is longer in the LMC and reaches smaller period ratios than in the SMC. Asimple test shows how the Petersen diagram is sensitive to the chemical composi-tion of the stars. We divided our sample of RRd stars into two groups – with period ol. 66
Fig. 4. Spatial distribution of double-mode RR Lyr stars in the OGLE fields toward the MagellanicClouds.
Upper panel presents RRd stars with period ratios P / P F > . Lower panel showspositions of RRd stars with P / P F ≤ . A. A. ratios above and below 0.744 – and we checked the spatial distribution of bothgroups. The result is displayed in Fig. 4. The P / P F ≤ .
744 group is almost ab-sent in the SMC, while in the LMC both groups have clearly different distributions,reflecting different metal abundance of these stars. RRd stars with P / P F ≤ . et al. (2016, in preparation).Anomalous RRd variables are characterized not only by different ratios of periodsin comparison to “classical” RRd stars, but also by different amplitude ratios (inthe anomalous RRd stars the fundamental mode usually dominates) and differentlight curve morphology of the fundamental mode component. Also, anomalousRRd stars usually show modulations of the pulsation amplitudes, in other words –the Blazhko effect. First RRd stars exhibiting Blazhko modulation were recentlydiscovered in the Galactic bulge (Soszy´nski et al. et al. In contrast to other types of classical pulsating stars (classical Cepheids – e.g. ,Udalski et al. et al. (2003) discovered three RR Lyr stars in the LMC which show additionaleclipsing variability superimposed on the pulsation light curves. However, it is notclear whether the pulsating stars are components of the binary systems, or theseare physically unrelated blends. Soszy´nski et al. (2009) found one more RR Lyrstar with eclipsing modulation. Very similar object – OGLE-BLG-RRLYR-02792– detected by Soszy´nski et al. (2011) in the Galactic bulge was spectroscopicallystudied by the Araucaria project (Pietrzy´nski et al. ⊙ . It turned outthat OGLE-BLG-RRLYR-02792 is a prototype of a new class of internal variables– binary evolution pulsators – that mimic properties and behavior of RR Lyr stars.In the present investigation, we report the discovery of one more candidatefor an RR Lyr star with eclipsing-like modulation – OGLE-LMC-RRLYR-30844– and we confirm the four previously announced objects of this kind (Soszy´nski etal. I -band light curves of all five stars.Based solely on the photometric data it cannot be judged whether these ob-jects are real binary systems with a pulsating star as one of the components or theeclipsing binaries are not related to the RR Lyr stars and are just optical blends. Itshould be noted that the orbital periods of our candidates are very short compared ol. 66 Fig. 5. OGLE-IV I -band light curves of RR Lyr stars showing eclipsing variability. Left panels showthe original photometric data folded with the pulsation periods.
Right panels show the eclipsing lightcurves after subtracting the RR Lyr component. The ranges of magnitudes are the same in each pairof the panels. A. A. to the values expected for horizontal branch stars which in the previous stage oftheir evolution were located at the tip of the red giant branch. Recently, Hajdu etal. (2015) conducted a search for binary RR Lyr stars in the Galactic bulge usingthe OGLE collection (Soszy´nski et al. P orb = .
24 d) indicatesthat at least some of the stars shown in Fig. 5 may be binary evolution pulsators– stars that transferred most of their mass to their companions and currently theycross the pulsation instability strip in their fast evolution toward the helium whitedwarf branch. In particular, OGLE-LMC-RRLYR-03541 ( P orb = .
23 d) hasvery similar properties (period, light curve shape) to OGLE-BLG-RRLYR-02792.In turn, OGLE-LMC-RRLYR-10752 exhibits a monotonic decrease of the pul-sation period, just as it is expected for fast-evolving binary evolution pulsators(Pietrzy´nski et al. − . ± .
01 s/yr.
7. Conclusions
We presented the OGLE collection of over 45 000 RR Lyr stars in the Magel-lanic System. Our sample contains, in fact, the vast majority of all RR Lyr vari-ables in the Magellanic Clouds. A comparison of the OGLE-IV set to the previouseditions of the OGLE collection of variable stars (Soszy´nski et al.
Acknowledgements.
We would like to thank Profs. M. Kubiak and G. Pietrzy´n-ski, former members of the OGLE team, for their contribution to the collection ofthe OGLE photometric data over the past years. We are grateful to Z. Kołaczkowskiand A. Schwarzenberg-Czerny for providing software used in this study.This work has been supported by the Polish Ministry of Science and Higher Ed- ol. 66
Alard, C., and Lupton, R.H. 1998,
ApJ , , 325.Alcock, C., et al. http://vizier.u-strasbg.fr/viz-bin/VizieR?-source=II/247 .Bhardwaj, A., Kanbur, S.M., Singh, H.P., and Ngeow, C.-C. 2014, MNRAS , , 2655.Bica, E., Bonatto, C., Dutra, C.M., and Santos, J.F.C. 2008, MNRAS , , 678.Chen, B.-Q., Jiang, B.-W., and Yang, M. 2013, Research in Astronomy and Astrophysics , , 290.Deb, S. and Singh, H.P. 2014, MNRAS , , 2440.Deb, S., Singh, H.P., Kumar, S., and Kanbur, S.M. 2015, MNRAS , , 2768.Feast, M.W., Abedigamba, O. P., and Whitelock, P.A. 2010, MNRAS , , L76.Gavrilchenko, T., Klein, C.R., Bloom, J.S., and Richards, J.W. 2014, MNRAS , , 715.Hajdu, G., Catelan, M., Jurcsik, J., Dékány, I., Drake, A. J., and Marquette, J.-B. 2015, MNRAS , ,L113.Haschke, R., Grebel, E.K., and Duffau, S. 2011, AJ , , 158.Haschke, R., Grebel, E.K., Duffau, S., and Jin, S. 2012a, AJ , , 48.Haschke, R., Grebel, E.K., and Duffau, S. 2012b, AJ , , 106.Kapakos, E., Hatzidimitriou, D., and Soszy´nski, I. 2011, MNRAS , , 1366.Kapakos, E., and Hatzidimitriou, D. 2012, MNRAS , , 2063.Kim, D.-W., Protopapas, P., Bailer-Jones, C.A.L., Byun, Y.-I., Chang, S.-W., Marquette, J.-B., andShin, M.-S. 2014, A&A , , A43.Kim, D.-W., and Bailer-Jones, C.A.L. 2016, A&A , , A18.Long, J.P., El Karoui, N., Rice, J.A., Richards, J.W., and Bloom, J.S. 2012, PASP , , 280.Moretti, M.I., et al. MNRAS , , 2702.Muraveva, T., et al. ApJ , , 127.Pejcha, O., and Stanek, K.Z. 2009, ApJ , , 1730.Pietrzy´nski, G., et al. Nature , , 75.Popielski, B.L., Dziembowski, W.A., and Cassisi, S. 2000, Acta Astron. , , 491.Ripepi, V., Moretti, M.I., Clementini, G., Marconi, M., Cioni, M.R., Marquette, J.B., and Tisserand, P.2012, Astrophysics and Space Science , , 51.Simon, N.R., and Lee, A.S. 1981, ApJ , , 291.Smolec, R., et al. MNRAS , , 3756.Soszy´nski, I., et al. Acta Astron. , , 369.Soszy´nski, I., et al. Acta Astron. , , 93.Soszy´nski, I., et al. Acta Astron. , , 1.Soszy´nski, I., Udalski, A., Szyma´nski, M.K., Kubiak, J., Pietrzy´nski, G., Wyrzykowski, Ł.,Ulaczyk, K., and Poleski, R. 2010, Acta Astron. , , 165.Soszy´nski, I., et al. Acta Astron. , , 1.Soszy´nski, I., et al. Acta Astron. , , 177.Subramaniam, A., and Subramanian, S. 2009, A&A , , L9.Subramanian, S., and Subramaniam, A. 2012, ApJ , , 128.Udalski, A., Szyma´nski, M.K., and Szyma´nski, G. 2015a, Acta Astron. , , 1.Udalski, A., et al. Acta Astron. , , 341.Wagner-Kaiser, R., and Sarajedini, A. 2013, MNRAS , , 1565.Wo´zniak, P.R. 2000, Acta Astron. ,50