The catalog of short periods stars from the ”Pi of the Sky” data A. Majczynaa, M. Nalez˙ytyb, M. Biskupc, G. Wrochnaa, M. Sokołowskia, K. Nawrockia, K. Małekd, L. Mankiewiczd, 8 0 L.W. Piotrowskic 0 2 aTheAndrzejSołtanInstitute forNuclearStudy, Hoz˙a69,Warsaw,Poland; n a bWarsawUniversity Observatory, Al.Ujazdowskie4,Warsaw,Poland; J cFacultyofPhysicsWarsawUniversity, Hoz˙a69,Warsaw,Poland; 7 1 dCenterforTheoretical Physics,PolishAcademyofScience, Al.Lotniko´w32/46,Warsaw, ] Poland; h p - o r Abstract t s a Based onthedata from the ”Piofthe Sky”project wemadeacatalog ofthevariable stars [ withperiodsfrom0.1to10days.Weuseddatacollectedduringaperiodoftwoyears(2004 1 and 2005) and classified 725 variable stars. Most of the stars in our catalog are eclipsing v 9 binaries - 464 (about 64%), while the number of pulsating stars is 125 (about 17%). Our 5 classification is based on the shape of the light curve, as in the GCVS catalog. However, 6 2 some stars in our catalog were classified as of different type than in the GCVS catalog. . Wehavefound periods for15starspresent intheGCVScatalog withpreviously unknown 1 0 period. 8 0 : Keywords: Stars:variables: catalog v i X r a 1 Introduction The ”Pi of the Sky” (Burd et al. 2005 [2], and the project’s web page [5]) was de- signedandbuiltbyateamfromtheAndrzejSołtanInstituteforNuclearStudy,Fac- ulty of Physics Universityof Warsaw, Warsaw Universityof Technology and Cen- ter for Theoretical Physics of the Polish Academy of Science. This project applies small telescopes to observe bright objects according to the idea proposed by prof. Bohdan Paczyn´ski (Paczyn´ski 1997 [8]). In 2004 thanks to Grzegorz Pojman´ski from Warsaw University Observatory, the ”Pi of the Sky” project was installed in ASAS (Pojman´skiG., 1997 [9]) domein Las Campanas Observatory (operated by theCarnegieInstitutionofWashington). PreprintsubmittedtoNewAstronomy 2February2008 In this work we report on observations from the first phase of the project dat- ing from July 2004 to December 2005. At this time the telescope was equipped with 2048×2062 CCD cameras with 15×15µm pixels and 12 e− readout noise using Carl-Zeiss telephoto lenses with d = f/1.4. The robotic telescope consists of two cameras installed on one mount. Both cameras observe the same 33◦×33◦ field.ThetelescopelocatedatLasCampanasObservatorycouldobservestarswith ◦ ◦ RA=<0;24> hr and Dec=<-89.8 ;+36 >. We do not useany filter except for IR-cut onein ordertomaximizethelimitingmagnitude. The main goal of the ”Pi of the Sky” is searching for the gamma ray burst optical afterglows so wheneverit is possible,cameras track theSwift orINTEGRAL field of view. In the remaining time the telescope observes interesting objects from a specially prepared list, so strong observational selection effects can be present in thedata. 2 Data Reduction Thedatareductionwasmadeautomaticallyusingsoftwareandscriptsadaptedfrom theASASprojectorcreatedspeciallyforthe”PioftheSky”.Dataweredividedin two streams, the first obtained directly from 10 seconds exposures and the second from 200 s exposures obtained by co-adding of 20 images. Due to the readout time the time resolution is 12 s and 240 s respectively. During each night a large amount of data is produced (about 3GB/hr), so only the results are saved in our database – raw images are deleted after one week. The photometry is made with different apertures – a small one for faint stars and a large one for bright objects. System design and observational strategy determine limiting magnitude to about 10–11 mag for 10sec exposure and ∼ 11–12 mag for 200 s (Burd et al. 2005 [2]). As the ”Pi of the Sky” does not use any filter, except for the IR-cut one, transition from the instrumental magnitude to the V filter magnitude is a source of a sys- tematic error. Unfortunately, value of this error is different for different stars. The formal photometry error described as rms is equal to ∼ 0.07 mag, but dispersion of the observationalpoints in the light curveis about 0.1 mag for stars fainter than 9 mag. A large correction is needed for different positions of stars on the CCD duetostrongvinietingand opticaldistorsionsoflenses. To avoidthisproblemone could use data taken from one field only, but then the number of data points drops dramatically.Avisualinspectionwashelpfultofindcompromisebetweenthenum- berofpointsandmeasurementquality. Thestar identificationin the”Pi oftheSky” databaseis based on comparison with stars from the Tycho-2 catalog (ESA 1997 [3], Høg 1997 [4]). Matching is based mainly on coordinates. Only a crude check of magnitude is performed because 2 of the filter correction problem mentioned above. The identification procedure as- sumesthatidentificationispositiveifaninvestigatedstariscloserthan2acrminto a star in the Tycho-2 catalog. An estimated error of astrometry is about 0.5 acrmin (Biskup2007 [1]). 3 Searching forvariability Wecheckedourdatatofindallvariablestarsanddeterminetheirvariabilityperiods. All measurements for 925201 stars were analyzed. We used the AoV algorithm (Schwarzenberg-Czerny1989[11])todetermineperiods,andrejectedstarswiththe statisticQ largerthan50.0.Wecheckedperiodsintherangefrom0.1to10daysfor stars with a number of observationalpointslarger than 200. Next, somestars were rejected during the visual inspection, so only 725 stars were classified as variable stars. The details of this procedure are described in Biskup (2007) [1]. In the next stepwe determinedvariabilitytypeofeach star. Starsweredividedintwogroups(Biskup2007[1])-starsexistingintheASAScat- alog(Pojman´ski1997[9])andtheseexistingonlyintheGCVScatalog(Kholopov et al. 1985 [6], Kholopovet al. 1992 [7]). Theautomaticprocedure ofa period de- termination returned half period instead of the full one for a significant number of stars (about 26%). For this reason each period was checked visually and corrected ifnecessary. Table1 Starsandperiodswhichweredetermined forthefirsttimefromourdata. ID RA Dec P[d] Piclass PGC GCVSclass othername 000360+1208.8 00:03:60 12:08:46 0.1701 DSCT no DSCTC NNPeg 001231+1433.8 00:12:31 14:33:51 1.8178: var no RS LNPeg 023602+0625.8 02:36:02 06:25:49 0.2079 DSCT no DSCT DXCet 052760+1254.8 05:27:60 12:54:45 0.3793: var no EB V1371Ori 104508+1620.1 10:45:08 16:20:07 0.2043 EW no EW EXLeo 114157-2423.2 11:41:57 -24:23:10 0.1363 DSCT no DSCTC VYCrt 120920-2759.3 12:09:20 -27:59:18 0.2923 EB no EB QYHya 124420-0840.3 12:44:20 -08:40:16 0.1167 EA no EA/D HWVir 132654-0555.7 13:26:54 -05:55:40 0.4923 EB: no EB LUVir 141420-1521.2 14:14:20 -15:21:11 0.2984 DSCT/BY no BY MVVir 141742-2149.6 14:17:42 -21:49:37 0.1539 DSCT: no DSCTC MXVir 143205-2742.7 14:32:05 -27:42:40 0.9123 EB no EB V0356Hya 150401-2803.7 15:04:01 -28:03:43 0.1466 DSCT no DSCT HYLib 173737-4048.8 17:37:37 -40:48:48 3.3864 DCEP no DCEPS V0950Sco 234535+2528.5 23:45:35 25:28:31 0.5790 EW no EW V0357Peg In Table 1 we present a list of 15 stars with previously unknown periods, which were determined from our data. These stars are present in the GCVS catalog, but 3 withoutperiod determinations. The classification is based on the shape of the light curve analogically to the pro- cedure in the GCVS catalog and in some cases based on the additional informa- tion about a spectral type. Our strategy is different from that in the ASAS catalog, where the classification is based on a decompositionof the light curve (Pojman´ski 2002[10]).Variabilitytypesweredeterminedsimplybyavisualinspectionoflight curves. Symbols denoting variability types used in our classification are summarized in Table 2. In addition, we have introduced symbols describing a situation when the classification is ambiguous: E - eclipsing binary, var - variable star, and ”:” uncer- tain. Table2 Differenttypesofvariability described inourcatalog ofvariablestars. Symbol Nameofprototype Period(days) Amplitude(inVfilter) Eclipsingbinaries EA Algol 0.2–1000 <several EB b Lyrae >1.0 <2 EW WUrsaMaioris <1.0 <0.8 Pulsatingstars RRAB RRLyrae 0.3–1.2 0.5–2.0 RRC RRLyrae 0.2–0.5 <0.8 DCEPS d Cephei <7.0 <0.5 BCEP b Cephei 0.1–0.6 0.01–0.3 DSCT d Scuti 0.01–0.2 0.003–0.9 CW WVirginis 0.8–35.0 0.3–1.2 Other ACV Alpha2CVn 0.5–160.0 0.01–0.1 INT/IT — 1.0–10.0 <1.0 In case of the classification of eclipsing binaries b Lyrae type (EB) and W Ur- sae Majoris (EW) we used two criteria. First, we assumed that for EW stars the period should be shorter than 1.2d. Second, the secondary minimum of the light curve should be deeper than 1/3 depth of the primary minimum. If at least one of theseconditionshadnot been fulfilled,such starwas denotedas EB/EW. 4 Fig.1.(a)thelightcurveoftheSTTaustarofDCEPtype,withtheperiodP=4.0259d;(b) phasedlightcurve Fig.2.PhasedlightcurveoftheRVCrvstarofEWtypewiththeperiodP=0.7473d 4 Conclusions 4.1 Eclipsingbinaries In Table 3 we summarized the number of stars with a different type of variability. The most numerous group of stars is of the EW type. This is in disagreement with 5 theGCVScatalog,wherethenumberofEWstarsisabout4timessmallerthanthe number of EA stars. On the other hand, the number of EB type stars is almost as large as thenumberofEWtypestars much likein theGCVS catalog. Such effects are artificial, resultingfrom assumptionsmadeduring searching forthevariability. Weinvestigatedonlyvariablestarswithperiodsintherangefrom0.1dto10days, so for that reason there is an excess of stars with short periods in comparison with theGCVScatalog. Table3 Numberofstarsofeachtypefromtheproject”PioftheSky” Type Numberofstars Type Numberofstars EA 43 RR 2 EA: 4 RRAB 36 EB 83 RRC 11 EB: 10 RRC/DSCT 3 EW 163 RR/DCEP 1 EW: 50 DCEP 18 EA/EB 40 DCEPS 4 EB/EW 35 BCEP 1 E 10 DSCT 48 E: 26 BCEP/DSCT 1 EW/RR 1 DSCT/BY 1 EW/RRC 5 CW 5 EW/DSCT 12 ACV 3 INT/IT 1 CW/DCEP 1 EW/DSCT/RRC 1 var 73 var: 33 Such discrepancy disappears when we comapare our results with the ASAS cata- log.NotethatPojman´ski(1997[9])useddifferentclassificationcriteriathaninour work, so there is no simple relation between e.g. EW and EC stars, and a compar- ison of results from these two catalogs is difficult. However, it is possible to make some general conclusions. Table 2 of Pojman´ski (2002) [10] shows that most of binary stars are contact ones - EC type, where both stars fill their Roche lobe. In ourcatalog thesituationissimilar,themostnumerousareEWtype(WUMatype) binary stars. Less numerous in our catalog are EB and EA types stars and a simi- lar trend appears in the ASAS catalog for ESD and ED stars as well. One should keep inminddifferences inclassificationcriteriaintheASASand Pi catalogs,and strongselectioneffects inthe”Pi oftheSky”data. 4.2 Pulsatingstars Discrepancy between the number of different types of pulsating stars in the ”Pi of the Sky” and GCVS catalogs can not be explained by the simple fact that we 6 Fig.3.PhasedlightcurveoftheRRLeo,RRABtypestarwiththeperiodP=0.4520d investigatedonlystarswithperiodsshorterthan10days.Allpulsatingstarspresent inourcataloghaveperiodsmuchshorterthan10days.Forexample,theratioofthe numberofRRABstarstothenumberofallstarsinourcatalogis0.05,whileinthe GCVS catalog it is equal to 0.11. Comparing the ratio of numbers of RRAB and DSCTtypestars,weobtainedthefollowingvalues:ASAS1.7(basedonPojman´ski 2002 [10]), ”Pi of the Sky” about 0.75, and GCVS 39.6. Similar situation occurs for other types of pulsating stars as well. The main difference between the ”Pi of the Sky” or ASAS and GCVS catalogs is the magnitude range. In the GCVS catalog there are informations about stars with -1.4–20 mag, whilethe magnitude rangefor”Pi oftheSky”is5–12mag andforASAS8.5–15mag.Weinterpreted differences in the number of each type of stars in these 3 catalogs as an effect of a different magnitude range in theses projects and effect of a specific observational strategyinthe”Pi oftheSky”project. 5 Summary and future perspective We presented the catalog of 725 variable stars from the first phase of the ”Pi of the Sky” project. The catalog contains stars with periods ranging beetwen 0.1 and 10 days.Thetypesofvariabilityweredetermined throughavisualinspection. Most of the variable stars in our catalog are eclipsing binaries, W UMa type. Among the pulsating stars most of them are d Scuti type stars (48), but RRAB starsareonlyslightlylessnumerous(36).Both typeofstarshaveratherlargevari- abilityamplitude,so are easy to detect. We determineaccurate periods for15 stars from theGCVS catalog withperiodsunknownsofar. The catalog and whole data base of measurements are available at the ”Pi of the Sky”websitehttp://grb.fuw.edu.pl/pi/index.html#star_catalog 7 Since the primary goal of the ”Pi of the Sky” is searching for the GRB prompt optical emission, the algorithm has been optimized for the flash recognition. We workonimprovingthephotometryandweplantodevelopanautomaticprocedure for a classification of variable stars. This will be necessary for the final version of the ”Pi of the Sky” project where the amount of data will increase by almost two orders ofmagnitude. Acknowledgments. We are very grateful to Grzegorz Pojman´ski for access to the ASAS domeand manypractical advices. We would like to thank the staff of the Las Campanas Observatory for their help duringtheinstallationand maintenanceofourtelescope. This work was financed by the Polish Ministry of Science in 2005-2007 as a re- search project. This work is also supported by the Polish Ministry of Science and HigherEducattiongrant No.NN203406133(AM, MN). References [1] Biskup, M. 2007, ”Variable stars search in the Pi of the Sky experiment” Master’s thesis, FacultyofPhysicsWarsawUniversity [2] Burd,A.etal.2005,NewAstronomy10,409 [3] ESA1997,TheHipparcos andTychoCatalogues, ESASP-1200 [4] HøgE.,1997Proceedings ofESASymposium’Hipparcos -Venice’97’, ESA-SP402 [5] http://grb.fuw.edu.pl [6] Kholopov, P.N. et al. 1985-1988, General Catalogue of Variable Stars 4rd ed. vols. I-III,Nauka,Moscow [7] Kholopov, P.N. et al.1992, General Catalogue of Variable Stars 4rd ed. vol. IV, Bull. Inf.CDS40,15 [8] Paczyn´ski B.,1997 TheFutureofMassiveVariability Searches, inVariable Starsand theAstrophysical ReturnsoftheMicrolensing Surveys.EditedbyRogerFerlet,Jean- PierreMaillardandBrigitteRaban.Cedex,France:EditionsFrontieres, 1997,p.357. [9] Pojman´ski;G.1997, ActaAstronomica47,467 [10] Pojman´skiG.2002,ActaAstronomica52,397 [11] Schwarzenberg-Czerny, A.1989, MNRAS241,153 8