Table Of Content

Mon.Not.R.Astron.Soc.000,000–000 (0000) Printed13January2011 (MNLATEXstylefilev2.2) Cepheid investigations using the Kepler space telescope R. Szabo´1⋆, L. Szabados1, C.-C. Ngeow2, R. Smolec3, A. Derekas1,4†, 1 P. Moskalik5, J. Nuspl1, H. Lehmann6, G. Fu˝r´esz7, J. Molenda-Z˙akowicz8, 1 0 S. T. Bryson9, A. A. Henden10, D. W. Kurtz11, D. Stello12, J. M. Nemec13, 2 J. M. Benko˝1, L. Berdnikov14,15, H. Bruntt16, N. R. Evans17, N. A. Gorynya18, n a E. N. Pastukhova18, R. J. Simcoe14, J. E. Grindlay19, E. J. Los19, A. Doane19, J S. G. Laycock19, D. J. Mink17, G. Champine19, A. Sliski19, G. Handler3, 2 1 L. L. Kiss1, Z. Kolla´th1, J. Kova´cs20, J. Christensen-Dalsgaard21, H. Kjeldsen21, ] C. Allen22, S. E. Thompson23, J. Van Cleve23 R 1Konkoly Observatory of the Hungarian Academy of Sciences, Konkoly Thege Miklo´s u´t 15-17, H-1121 Budapest, Hungary S 2Graduate Institute of Astronomy, National Central University,Jhongli City, Taoyuan County 32001, Taiwan . 3Institut fu¨r Astronomie, Universita¨t Wien, Tu¨rkenschanzstrasse 17, A-1180 Wien, Austria h 4Department of Astronomy, E¨otv¨os University, Budapest, Hungary p 5Copernicus Astronomical Center, ul. Bartycka18, 00-716, Warsaw, Poland - o 6Thu¨ringer Landessternwarte Tautenburg, Karl-Schwarzschild-Observatorium, 07778 Tautenburg, Germany r 7Harvard-Smithsonian Centerfor Astrophysics, MS 20, 60 Garden Street, Cambridge, MA 02138, USA st 8Astronomical Institute, Universityof Wroc law, ul. Kopernika 11, 51-622, Wroc law, Poland a 9NASA Ames Research Center, Moffett Field, CA 94035, USA [ 10American Association of Variable Star Observers,Cambridge, MA 02138, USA 11Jeremiah Horrocks Institute of Astrophysics, Universityof Central Lancashire, Preston PR1 2HE, UK 1 12Sydney Institute for Astronomy, School of Physics, The University of Sydney, Sydney, 2006 NSW, Australia v 13Department of Physics & Astronomy, Camosun College, Victoria, British Columbia, V8P 5J2, Canada 3 14Sternberg Astronomical Institute, Moscow University,Moscow, Russia 4 15Isaac Newton Institute of Chile, Moscow Branch, UniversitetskijPr. 13, Moscow 119992, Russia 4 16LESIA, UMR 8109, Observatoire de Paris, 92195 Meudon Cedex, France 2 17Smithsonian Astrophysical Observatory, MS4, 60 Garden Street, Cambridge, MA 02138, USA 1. 18Institute of Astronomy, Russian Academy of Sciences, 48 Pyatnitskaya Street, Moscow, 109017, Russia 0 19Harvard College Observatory, 60 Garden Street, Cambridge, MA 02138, USA 1 20Gothard Astrophysical Observatory, H-9707, Szombathely, Hungary 1 21Department of Physics and Astronomy, Aarhus University,DK-8000 Aarhus C, Denmark : 22Orbital SciencesCorporation/NASA Ames Research Center, Moffett Field, CA 94035 v 23SETI Institute/NASA Ames Research Center, MoffettField, CA 94035 i X r a Accepted 2011January12,Received2010December 31 ABSTRACT We report results of initial work done on selected candidate Cepheids to be observed with the Kepler space telescope. Priorto the launch40 candidates were selected from previous surveys and databases. The analysis of the first 322 days of Kepler pho- tometry,andrecentground-basedfollow-upmulticolourphotometryandspectroscopy allowed us to confirm that one of these stars, V1154Cyg (KIC7548061), is indeed a 4.9-dCepheid.Usingthe phaselagmethodweshowthatthisstarpulsatesinthe fun- damentalmode. New radialvelocity data are consistentwith previousmeasurements, suggestingthatalong-periodbinarycomponentisunlikely.No evidenceis seeninthe ultra-precise, nearly uninterrupted Kepler photometry for nonradial or stochastically excited modes at the micromagnitude level. The other candidates are not Cepheids but an interesting mix of possible spotted stars, eclipsing systems and flare stars. Key words: stars: oscillations – stars: variables: Cepheids 2 R. Szab´o, L. Szabados, C.C. Ngeow, et al. 1 INTRODUCTION 0 The Kepler1 space mission is designed to detect Earth- 1 like planets around solar-type stars with the transit method (Borucki et al. 2010) by monitoring continuously 2 over 150000 stars with an unprecedented photometric pre- V1154 Cyg cision. The lifetime of at least 3.5 years and the quasi- g g 3 continuous observations make Kepler an ideal tool to mea- o l sure stellar photometric variability with a precision that 4 is unachievable from the ground (see, e.g. Gilliland et al. 5 2010a). Ground-based photometric observations of Cepheids 6 usually consist of a few (typically one or two) observa- tions per night. Until now it has not been possible to ad- 7 14000 12000 10000 8000 6000 4000 2000 equately cover many consecutive pulsational cycles with Teff ground-based photometry. Ground-based studies of several Figure1.SelectionofCepheidcandidatesbasedonKeplerInput types of variable stars – such as δSct stars, DOV, DBV Catalog effective temperature and logg values (small squares). and DAV white dwarfs, sdBV stars and roAp stars – have Theplotcontains5987kasctargets(redplussigns);theselected benefited from multisite photometric observations. But be- Cepheidcandidatesareshownbyfull(blue)squares.V1154Cyg, cause of their longer period, Cepheids have not been the theonlyconfirmedCepheidsisdenotedbyalarge(blue)square. targetsofsuchconcertedefforts,withthenotableexception ThelinearCepheidinstabilitystripisdenoted bydashedlines. of V473Lyr(Burki et al. 1986). Space-based Cepheid observations were conducted ear- lier with thestar trackeroftheWIREsatellite andtheSo- candidates for the survey phase. First, all known, pos- lar Mass Ejection Imager (SMEI) instrument on board the sible or suspected variable stars in the relevant pul- Coriolis satellite (Bruntt et al. 2008; Spreckley & Stevens sational period range were selected from all available 2008; Berdnikov 2010). The lengths of these data sets (∼ databases containing variability information, such as 1000−1600d) are comparable to the nominal lifetime of GCVS (Samuset al. 2002), ASAS North (Pigulski et al. the Kepler Mission. The WIRE and SMEI data of Polaris 2009), ROTSE (Akerlof et al. 2000) and HAT catalogs (αUMi,V =2.005) confirmedthatthepulsationamplitude (Hartman et al. 2004). Light curve shapes and the logP of Polaris has been increasing again, after a long period of vs. J −H diagram (Pojmanski & Maciejewski 2004) were decrease. Kepler is capable of delivering comparably accu- also utilized for further selection. Finally, stars with close ratephotometricobservationsforthemuchfainterCepheid, (bright)companionswereexcluded.Thisprocedureresulted V1154Cyg (V =9.19). inalistof26stars.Wenotethatthesecatalogsarenotcom- The Kepler Asteroseismic Science Consortium kasc2 pletein coverage of thefield,norin therelevant magnitude wassetuptoexploittheKeplerdataforstudyingstellarpul- range. sations. kasc working group number 7 (WG7) is dedicated Asasecondapproach,wesearchedforstarslyinginside to the investigation of Cepheids. In compliance with one of orclosetotheCepheidinstabilitystripbasedontheKepler the original kasc proposals submitted before the launch of Input Catalog (KIC)3 effective temperature and logg val- the space telescope, we searched for Cepheids among a list ues.StarsfainterthanKp=16.0magwereexcluded,where of 40 targets, including the only previously known Cepheid KpdenotestheKeplermagnitudesystem(seeSection3.for V1154Cyg (KIC7548061) in the field. In this paper we de- moredetails on theKepler magnitudes). Candidateswith a scribethefirstresultsfromKeplerobservationsofaCepheid contaminationindex(CI)largerthan0.5werealsoomitted, and a couple of not confirmed Cepheids, complemented by whereCI=0 meansthat all thefluxin theaperturecomes extendedground-based follow-up observations. fromthetargetandincaseofCI=1thefluxentirelycomes fromsurroundingsources.Thisresultedin14additionaltar- getsasshowninFig.1.ThelinearCepheidinstabilitystrip, denoted by dashed lines in the figure, was calculated with 2 TARGET SELECTION theFlorida-Budapest code (Szab´o et al. 2007). We note that the process followed to derive the KIC Being supergiants, Cepheids are rare and the number of CepheidsintheKeplerfixed105squaredegreefieldofview parameters was optimized to find main-sequencestars with high probability and to distinguish them from cool giants (FOV) is expected to be low. Due to telemetry bandwidth constraints,Keplercannotobserveallstarsinthefield.The (Batalha et al. 2010). The precision of the parameters is therefore limited for our purposes. This is illustrated in first 10 months of operation (observing quarters Q0–Q4) Fig.1 by the fact the KIC parameters of V1154Cyg (the were therefore dedicated to a survey phase to find the best only previously known Cepheid in the field) put it outside candidates,whichwouldthenbeobservedduringtherestof thecomputed instability strip by several hundredKelvins. themission. Consequently, we proposed to observe the above men- We used two different approaches to find Cepheid tioned 40 stars by Kepler during the‘survey period’. 1 http://kepler.nasa.gov 2 http://astro.phys.au.dk/KASC/ 3 http://archive.stsci.edu/kepler Cepheid investigations using the Kepler space telescope 3 Table1.MainpropertiesoftheobservedKeplerCepheidandCepheidcandidates.Theuncertaintyoftheperiodisgiveninparentheses andreferstothelastdigit(s)oftheperiod.Q3.1meansthefirstmonthofQ3.Seetextforthedefinitionofthecontamination index. KICID Othername α δ Kp Period Contam. Runs J2000 J2000 mag days Index 7548061 V1154 Cyg 19 48 15.45 +43 07 36.77 8.771 4.925454(1) 0.036 LC: Q01234 SC: Q1 2968811 ASAS190148+3807.0 190148.21 +380701.99 13.469 14.8854(12) 0.047 LC:Q01234 6437385 ASAS192000+4149.1 192000.12 +414907.46 11.539 13.6047(7) 0.075 LC:Q01234 8022670 V2279Cyg 191854.46 +434925.82 12.471 4.12564(5) 0.189 LC:Q1234 12406908 ROTSE1 192344.99 +511611.75 12.354 13.3503(45) 0.015 LC:Q01SC:Q3.1 J192344.95+511611.8 tions are based on the same 6-s integrations which are summedtoformtheLCandSCdataonboard.Inthiswork we use BJD-corrected, raw LC data (Jenkins et al. 2010) spanningfromQ0toQ4,i.e.321.7dofquasi-continuousob- servations. Some of our targets (V1154Cyg among them) wereobservedinSCmodeaswell.Weexploitthisopportu- nity to compare LC and SC characteristics and investigate thefrequencyspectrumtoamuchhigherNyquistfrequency (733.4d−1 vs. 24.5d−1). The saturation limit is between Kp ≃ 11 − 12mag depending on the particular chip the star is observed; brighter than this, accurate photometry can be performed up to Kp ≃ 7mag with judiciously designed apertures (Szab´oet al. 2010). SinceV1154Cyg is much brighter than the saturation limit, it required special treatment, as can be seen in Fig.2, which shows a 50 pixel box centered on V1154Cyg. The plot was made using keplerffi4 written byM. Still. To illustrate some of the common characteristics of the data we show in Fig.3 the raw Kepler light curve of V1154Cyg after normalizing the raw flux counts and con- Figure2.VicinityofV1154CygfromoneoftheFullFrameIm- verting the fluxes to the magnitude scale. The small gaps ages.Thestarisslightlysaturatedasindicatedbyitsbrightness in thelight curvearedueto unplannedsafe modeandloss- andtheelongatedshape. of-fine-point events, as well as regular data downlink peri- ods.ThisLCdatasetspanningQ0–Q4datacontains14485 points. Fig. 4 shows a 33.5d segment (Q1) where SC data 3 KEPLER OBSERVATIONS are available for V1154Cyg containing 49032 data points. ThevaryingamplitudeseeninFig. 3isofinstrumental Kepler was launched on 2009 March 6 into a 372-d solar origin. It is a result of the small drift of the telescope, cou- orbit,andisobservinga105squaredegreeareaoftheskyin pled with different pixel sensitivities. In addition, different betweentheconstellationsofCygnusandLyra(Koch et al. aperture masks are assigned to the targets quarterly which 2010). After a 9.7d commissioning phase (Q0), the regular result in small changes in the measured flux. The most no- observations started on 2009 May 12. In order to ensure tableamplitudechangeisseentowardstheendofQ2,which optimal solar illumination of the solar panels, a 90degree was noted to be due to flux flowing outside of the optimal rollof thetelescope isperformed at theendofeach quarter aperture(bleeding),affectingthemeasuredbrightness.For- of its solar orbit. The first quarter lasted only for 33.5d tunately, in Q2 a larger mask was also downloaded besides (Q1), while subsequent quarters are all three months long. the standard optimal aperture assigned to this star which Ineach of thefourquartersannuallytheKeplertargets fall allowed us to investigate the variation of flux outside the on different CCDs. optimalaperture.Thisconfirmedthatthetotalfluxwasin- The Kepler magnitude system (Kp) refers to the wide deedcapturedwithinthelargeraperture,andhencethatthe passband (430−900nm) transmission of the telescope and star shows no intrinsic amplitude variation within the cur- detector system. Note that the Kepler magnitudes (Kp) rent accuracy of the data. Without further information we werederivedbeforethemissionlaunchandareonlyapprox- cannotchoosebetweenapossibleslightamplitudevariation imate values. Currently Kepler processing does not provide and instrumental effects in other quarters. calibrated Kepler magnitudes Kp (Kolenberg et al. 2010). Both long cadence (LC, 29.4min, Jenkins et al. 2010), and short cadence (SC, 58.9s, Gilliland et al. 2010b) observa- 4 http://keplergo.arc.nasa.gov/ContributedSoftwareKeplerFFI.shtml 4 R. Szab´o, L. Szabados, C.C. Ngeow, et al. 0.25 0.20 Q0 Q1 Q2 Q3 Q4 0.20 0.15 0.15 0.10 0.10 0.05 0.05 Kp Kp 0.00 ∆ 0.00 ∆ -0.05 -0.05 -0.10 -0.10 -0.15 -0.15 -0.20 -0.20 54950 55000 55050 55100 55150 55200 55250 54965 54970 54975 54980 54985 54990 54995 55000 BJD-2,400,000 BJD-2,400,000 Figure 3. Raw Kepler light curve of V1154Cyg. The varying Figure 4. Q1 SC Kepler light curve of V1154Cyg. It is almost amplitudeisinstrumentalinorigin;seetextformoredetails. indistinguishable from LC data taken during the same time in- terval.Notetheessentiallyuninterruptedsampling. 5.1 Multicolour photometry 4 TARGET CLASSIFICATION Ground-based multicolour observations provide additional After the release of the first Kepler light curves to the information, therefore complement the space-born photom- kasc it turned out that only a few stars showed Cepheid etrytakenin ‘whitelight’.Weusedthefollowing telescopes likevariability.ThepropertiesoftheseKeplerCepheidcan- to gather BVR I Johnson-Cousins magnitudes: c c didates are listed in Table1. Periods were determined for the Kepler LC light curves using the period04 program (i) Lulin One-meter Telescope (LOT) at Lulin Obser- (Lenz& Breger 2005). vatory (Taiwan)5: 1-m Cassegrain-telescope with PI1300B Apart from these candidates a number of other vari- CCD; able stars were included in the initial sample of 40 stars. (ii) SLT at Lulin Observatory (Taiwan): 0.4-m RC-tele- Based on their light curvesand theautomatic classification scope with Apogee U9000 CCD; (Blomme et al. 2010) we classify these as eclipsing binaries (iii) Tenagra telescope at Tenagra II Observatory6 (8 stars), ellipsoidal variables (6), delta Scuti stars (3), an (USA):robotic 0.8-m RC-telescope with SITe CCD; SPB star (1), long period variables (7) and stars with no (iv) SonoitaResearchObservatory(SRO,USA)7,0.35-m obviousvariations (7).While this sample is arich source of robotic telescope with a SBIG STL-1001E CCD. flaring, spotted, granulated stars, thedetailed investigation Observations were performed from September 2009 to ofthesenon-Cepheidsisoutofscopeofthispaper.Inaddi- August2010.Alloftheimagingdatawerereducedinastan- tion, we inspected all 6300 kasc light curves, but found no dardway(bias-subtracted,dark-subtractedandflat-fielded) Cepheids among them. using iraf8. Instrumental magnitudes for the stars in the Three originally proposed stars were not observed due imagesweremeasuredusingSExtractor(Bertin & Arnouts to technical reasons (position on the CCD chip, bright- 1996) with aperture photometry. For the SRO data a sep- ness/faintness, contamination orpixelnumberconstraints). arate photometric pipeline based on daophot was used, The large variety of non-Cepheids in our sample indi- whichincludedanalysisofanensembleofcomparison stars. cates that the spatial resolution and photometric accuracy Theinstrumentalmagnitudesweretransformedtothestan- of ASAS was not adequate to select such relatively faint dard magnitudes using Landolt standards (Landolt 2009). Cepheids.ItfurthershowsthattheKeplerInputCatalog is Times of observation were converted to heliocentric Julian clearly not optimal for our purpose. day. We gathered between 80 and 120 frames per target for eachpassband,excludingKIC12406908whichwasobserved only at TNG and hence had fewer images taken. We have found a small offset between SRO and other B data in the 5 GROUND-BASED FOLLOW UP case of V1154Cyg. Fortunately, data were taken in TNG and SRO very close in time (less than 1 minute) twice, al- To further confirm or disprove the Cepheid nature of our lowing us to correct the SRO data by the shift measured candidatesinTable1weemployedground-basedmulticolour photometry and in some cases spectroscopy. In the case of 5 http://www.lulin.ncu.edu.tw/english/index.htm V1154Cyg regular radial velocity observations and multi- 6 http://www.tenagraobservatories.com colour photometry were scheduled, which helped us to gain 7 http://www.sonoitaobservatories.org more information on the pulsation. In the following we de- 8 irafisdistributedbytheNational Optical AstronomyObser- scribe these ground-based follow-up observations and the vatories,whichareoperatedbytheAssociationofUniversitiesfor subsequent scrutinizing classification process on a star-by- Research in Astronomy, Inc., under cooperative agreement with star basis. theNationalScienceFoundation. Cepheid investigations using the Kepler space telescope 5 0.55 7.5 0.50 7.0 0.45 6.5 0.40 6.0 0.35 5.5 0.30 5.0 R21 0.25 φ21 4.5 0.20 4.0 0.15 KIC 2968811 3.5 KIC 6437385 0.10 KIC 12406908 3.0 0.05 V2279 Cyg 2.5 V1154 Cyg 0.00 2.0 0 5 10 15 20 25 30 0 5 10 15 20 25 30 P [days] P [days] 0.40 14.0 0.35 12.0 0.30 10.0 0.25 8.0 R31 0.20 φ31 6.0 0.15 4.0 0.10 0.05 2.0 0.00 0.0 0 5 10 15 20 25 30 0 5 10 15 20 25 30 P [days] P [days] Figure5.Fundamentalmode(redfilled)andfirstovertone(blueopensymbols)FourierparameterprogressionsinJohnsonV passband forGalacticCepheidsample.V1154CygandotherKeplerCepheidcandidatesarealsoshownbasedontheirnewground-basedJohnson V photometry. Error bars for V1154Cyg and V2279Cyg are comparable to, or smaller than the symbol size, so these were omitted from the plot.The Galactic Cepheid light curve parameters were compiled from Antonello&Lee (1981), Moffett&Barnes (1985), Antonello&Poretti (1986), Antonello,Poretti&Reduzzi (1990), Mantegazza &Poretti (1992), Poretti (1994), Antonello&Morelli (1996). Cepheid. We discuss our findings for individual candidates Table 2. Ground-based multicolour photometry of V1154Cyg andKeplerCepheidcandidates.Theentiretableisavailableonly below. electronically. KICID HJD Mag. Err. Filt. Obs. 5.2 Spectroscopy 02968811 2455103.7700 14.857 0.013 B SRO Spectroscopic observations were conducted to measure ra- 02968811 2455103.7809 14.823 0.013 B SRO 02968811 2455104.7544 14.797 0.011 B SRO dial velocity and derive stellar parameters. We obtained ... ... ... ... ... ... spectra of V1154Cyg with the Coude-Echelle spectrograph attachedtothe2-mtelescopeoftheThu¨ringerLandesstern- warte (TLS) Tautenburg. Spectra cover the region 470 − 740nm, the spectral resolution is R=33000, the exposure in these epochs: BSRO−BTNG =0.054. All the (corrected) time was 30min per spectrum. Spectra were reduced using photometric measurements are available online, while Ta- standard midas packages. The reduction included the re- ble2 shows the layout of thedata. moval of cosmic rays, bias and stray light subtraction, flat We decomposed the multicolour light curves to derive fielding, optimum order extraction, wavelength calibration the widely used Fourier parameters Ri1 and φi1, as de- using a Th-Ar lamp, normalization to the local continuum, fined by Simon & Lee (1981), which characterize the light and merging of the orders. Small instrumental shifts were curve shape. The Johnson V results are plotted in Fig.5 correctedbyanadditionalcalibrationusingalargernumber alongwiththeFourierparametersoffundamentalmode(red oftelluricO2 lines.Inaddition,onespectrumofV1154Cyg filled points) and first overtone (blue open circles) Galactic was taken on 2007 June 15 at the M. G. Fracastoro sta- Cepheids as a function of pulsational period. As all Galac- tion (SerraLaNave,Mt. Etna) oftheINAF - Osservatorio tic Cepheids follow the main Fourier progressions in Fig.5, Astrofisico di Catania (INAF-OACt). We used the 91-cm a star that lies outside the overall pattern is unlikely to be telescope and fresco, the fiber-fed reosc echelle spectro- 6 R. Szab´o, L. Szabados, C.C. Ngeow, et al. 0.10 0.05 KIC 2968811 0.08 0.04 KIC 6437385 0.06 0.03 0.04 0.02 0.02 0.01 Kp 0.00 Kp 0.00 ∆ ∆ -0.02 -0.01 -0.04 -0.02 -0.06 -0.03 -0.08 -0.04 -0.10 -0.05 55235 55240 55245 55250 55255 55260 55265 55270 55275 55235 55240 55245 55250 55255 55260 55265 55270 55275 BJD-2,400,000 BJD-2,400,000 0.30 0.15 KIC 8022670 V2279 Cyg KIC 12406908 0.20 0.10 0.10 0.05 Kp 0.00 Kp 0.00 ∆ ∆ -0.10 -0.05 -0.20 -0.10 -0.30 -0.15 55024 55026 55028 55030 55032 55034 55036 55090 55095 55100 55105 55110 55115 55120 55125 BJD-2,400,000 BJD-2,400,000 Figure 6.Representative partsof Keplerphotometry offour Cepheidcandidates (excluding theknown CepheidV1154Cyg). LC light curvesareplotted,exceptforKIC12406908whereonemonthofSCdataisplotted.The‘shoulders’onthedescendingbranchesindicate spottedstarsandtheflaresaretypicalforactivestars.TheselightcurvesindicatethatKIC8022670(V2279Cyg)isthemostpromising Cepheidcandidate ofthefour. graph which allowed us to obtain spectra in the range of first panel of Fig.7. We note that in the case of a Cepheid 4300−6800˚A with a resolution R=21000. in this period range the bump should be present on the For V2279Cyg we made a high-resolution spectrum ascending branch. By carefully examining the short bright- in a single-shot 1680 s exposure with the 1.5-m telescope eningseenintheKeplerlightcurveweexcludeinstrumental at the Fred Lawrence Whipple Observatory (Mt. Hopkins, or other external origin based on known artifacts discussed Arizona) using the Tillinghast Reflection Echelle Spectro- intheKeplerDataReleaseNotes9.Takingintoaccountthe graph(TRES)on2010September24.Thespectrumiscross- strong flare events (Fig.6), we conclude that this object is dispersed withawavelength rangeof3860−9100˚Aover51 definitely not a Cepheid. spectral orders. KIC6437385 = ASAS 192000+4149.1 (P = 13.6047d). This star shows a similar light curve to KIC2968811 with an amplitude of 0.09mag (Fig.6 upper 5.3 Remarks on individual candidates right panel). Within the uncertainty the Fourier parame- In the following we summarize the Kepler light curve char- ters R21 matches the progressions, but R31 does not. The phases are close to the phases of the bulk of the Galactic acteristics, the multicolour photometry, the Fourier param- Cepheids(Fig.5).TheB amplitudeissmallerthantheam- eters and the spectroscopic properties of the four Cepheid plitudemeasuredinV asshowninFig.7,thisisinconsistent candidates (see Table1), before moving to the previously with the Cepheid classification. The light variation shows known Cepheid (V1154Cyg) in Sect. 6. thesame shoulders on thedescendingbranch asin thecase KIC2968811 = ASAS 190148+3807.0 (P = ofKIC2968811,whileitsamplitudeandshapearechanging 14.8854d). In principle the light curve shape matches all throughout the more than 300d long Kepler observations. the Fourier parameter progressions within the uncertainty, In addition, several flares can be seen that are intrinsic to although in the case of φ21 and R31 it is a border line thestar.Oneisshown inFig.6.Thus,KIC6437385 ismost case (Fig.5).Theamplitudeis0.19mag intheKeplerpass- probably not a Cepheid. band.‘Shoulders’orbumpsappearonthedescendingbranch (Fig.6,upperleftpanel),buttheyarenotpresentatthebe- ginningoftheobservations.This effectis frequentlyseenin spotted stars. The same effect may cause the larger scatter of the folded ground-based multicolour light curves in the 9 http://archive.stsci.edu/kepler/data release.html Cepheid investigations using the Kepler space telescope 7 Figure 7. Phased BVRcIc Johnson-Cousins phase-folded light curves of three Cepheid candidates taken during the 2010 observing season: KIC2968811 (left panel), KIC6437385 (middle panel) and KIC8022670 (right panel). For clarity, error bars were plotted only fordatapointswitherrorslargerthan0.02mag. KIC12406908=ROTSE1J192344.95+511611.8 thephotometricsurvey byPigulski et al. (2009) resulted in (P = 13.3503d). Interestingly, this variable star exhibits usefuldataforfollowingtheperiodchangesofthisstar.Ad- verysimilarlightcurvecharacteristicstothepreviouslydis- ditionalphotometricdataarealsoavailablefromtheSuper- cussed objects (Fig.6, bottom right panel). It has an am- WASPpublicarchive10andtheScientificArchiveoftheOp- plitudeof0.3magin theKeplerpassband.Wehavetoofew tical Monitoring Camera (OMC) on board INTEGRAL11. data points from ground-based observations to put strong Both these archives contain data obtained in a single pho- constraintsontheFourierparameters,butdespitethelarge tometric band, and we used these data to investigate the uncertainties this object seems to be slightly off the main period behaviourof V2279Cyg. location expected for Cepheids (Fig.5). The ratio of I and TheKeplerlightcurveofV2279Cygseemsstable,with- c V amplitudes is 0.74, which is close to, but slightly larger, outanynoticeable light curvechanges (partof theQ2light thanthestandard0.6.TheSClightcurvesshowsmallshort- curve is plotted in Fig.6). We only see long term varia- duration flares at BJD2455106.1 and 2455110.6, and a tions similar in amplitude to what we noted for V1154 Cyg largeoneatBJD2455118.2whichshowedacomplex,multi- (Fig.3), which in this case might also just be instrumen- peakedmaximumlastingfor7h.Wenotethatafewoutliers tal effects. What makes V2279Cyg particularly suspicious were removed from the Kepler light curve, which however is the presence of many flares, one of them is clearly seen does not influence our conclusions. Based on the observed at BJD2455030.0 in Fig.6. The (almost) strictly periodic features we can safely exclude this star from our Cepheid variations and the value of the period is consistent with a sample. rotational modulation. KIC8022670 = V2279Cyg (P =4.12564d). Based Multicolour photometry can be decisive in solving this on the space and ground-based follow-up observations this classification problem because Cepheids have characteristic starisastrongCepheidcandidate.Wethereforelookedinto amplitude ratios. Our new photometric data suggest that the literature for a more rigorous assessment of whether V2279CygisnotaCepheid.Theamplitudeofitsbrightness Cepheid pulsation is the cause of the observed variability. variation in V is only slightly smaller than the amplitude Variability of this star was revealed by Dahlmark (2000) in the B band: the ratio is about 0.9 instead of the usual during a photographic search for variable stars, and in- valueof0.65−0.70 (Klagyivik & Szabados2009).Although dependently by Akerlof et al. (2000) as a result of the abrightbluecompanionstarisabletosuppresstheobserv- ROTSE project. While Akerlof et al. (2000) classified the able amplitude of the light variations in the B band, the star as a Cepheid with a period of 4.12298d, Dahlmark observed amplitude ratio is incompatible with the Cepheid (2000) interpreted it as an RS CVn type variable due to natureevenif thestar had a veryhot companion. R21,R31 its proximity to a known X-ray source detected by the and φ31 place it among the first overtone objects, but φ21 ROSAT mission (Voges et al. 1999). This object, ROTSE is very different from both fundamental and first overtone J191853.61+434930.0 = LD349, was finally designated progressions (Fig.5). as V2279Cyg among the variable stars (Kazarovets et al. In a next step we analyze the cycle-to-cycle variations 2003). in the periodicity using the so-called O−C diagram. The WefirstlookedatROTSE-Iphotometricdataretrieved O−C diagram constructed for the moments of brightness from the NSVS data base (Wo´zniak et al. 2004), but con- maximaisplottedinthetoppanelofFig.8.Thezeroepoch clude that this data set does not allow us to unambigu- was arbitrarily chosen at the first maximum in the Kepler ously classify the star. A dedicated photometric project to dataset. A least squares fit totheO−C residuals resulted select TypeII Cepheids among the ROTSE-I targets was performed by Schmidtet al. (2007). Based on their two- colour(V,R)photometrytheyconcludedthatV2279Cygis 10 http://www.wasp.le.ac.uk/public/index.php aprobableCepheid with aperiod of4.117d.More recently, 11 http://sdc.laeff.inta.es/omc/index.jsp 8 R. Szab´o, L. Szabados, C.C. Ngeow, et al. 200 1000 Na D Li Ca II H 800 150 1000 U U 600 U D 100 D D A A 400 A 500 50 200 0 0 0 3940 3960 3980 4000 5860 5880 5900 5920 5940 6660 6690 6720 6750 wavelength (Angstrom) wavelength (Angstrom) wavelength (Angstrom) Figure 9.ThetresspectrumofV2279Cyg.Fromlefttoright:zoom intoCaIIH(3968.5˚A),NaD(5890and5896˚A),Li(6708˚A). Table 3. Sample table for O−C residuals for V2279 Cyg. The entiretableisavailableonline. JD⊙ − E O−C W Reference 2400000 [d] 51474.9729 −843 −1.7218 2 NSVS 52626.3311 −564 −1.4177 1 Integral OMC 53051.2057 −461 −1.4842 3 Schmidtetal. 53368.6341 −384 −1.7304 1 Schmidtetal. 54265.3816 −191 −1.2317 1 SWASP 54645.2207 −75 0.0030 1 SWASP 54954.6214 0 0.0105 5 Kepler ... ... ... ... ... plotted three segments of the spectrum taken with the tres spectrograph in Fig.9 containing these lines: CaiiH (3968.5˚A),NaD (5890 and 5896˚A) and Li(6708˚A).While the NaD lines are normal, the characteristics of the two other lines do not support a Cepheid nature of V2279Cyg. TheLi(6708˚A)isneverseeninemissioninCepheids,while theCaemissionimplieschromosphericactivity,whichisalso Figure 8. Top panel: O−C diagram of V2279Cyg for the Ke- notaCepheidcharacteristic. Wefittedtheoreticaltemplate plerdata.Bottom panel:O−C diagramofV2279Cyginvolving spectra from the extensive spectral library of Munari et al. all photometric data. The point size corresponds to the weight (2005)tothespectrum anddeterminedthefollowing atmo- assignedtothemaxima. spheric parameters: Teff = 4900±200K, logg = 3.7±0.4, [M/H] = −1.2±0.4 and vsini = 40kms−1. The resulting parameters are also incompatible with a Cepheid variable, in the best fitting period of 4.125642d, which is somewhat but suggest a cool main-sequence star with moderate rota- longerthananyformerlypublishedvalueforV2279Cyg.We tion. thenincludedallground-basedobservationswecouldfindto Summarizing the previous subsections, we conclude construct a newO−C diagram using the ephemeris that all the candidates turned out not to be Cepheids ex- C =BJD(2454954.6109±0.0023)+(4.125642±0.000051)E. cept the already known Cepheid V1154Cyg, which we will Table3 contains the moments of maxima, the correspond- describe in detail in the following. ing epochs and the O−C residuals of the available obser- vations. We assigned different weights to data from differ- ent sources. These weights (between 1 and 5) correspond 6 V1154 Cyg THE ONLY KEPLER CEPHEID to the ‘goodness’ of the seasonal light curve, larger num- ber means better coverage and smaller scatter (for Kepler In this section we analyze both Kepler data and ground- data W=5). The period of V2279Cyg show large fluctua- basedfollow-up observationsof what isapparentlytheonly tions during the last decade (see the lower panel of Fig.8). Cepheid being observed in Kepler’s FOV. Our additional ground-based photometric data indicate a very recent change in the period that we will investigate 6.1 Observational data prior Kepler using future quartersof Kepler data. Wenote that the star hasahighcontaminationindex(Table1),indicatingthatthe Brightness variation of V1154Cyg was discovered by ∼18percentofthemeasuredfluxcomesformothersources. Strohmeier et al. (1963). Cepheid type variation and a pe- ThespectrumofV2279Cygofferstheclearest evidence riodicity somewhat shorter than 5d were obvious from that this star has been misclassified as a Cepheid. We have the photographic magnitudes leading to the discovery. Cepheid investigations using the Kepler space telescope 9 Figure 10. BVRcIc Johnson-Cousins phase-folded light curves Figure 11. Comparison of the extinction-corrected colours of of V1154Cyg taken during the 2010 observing season. Typical V1154Cyg(filledsquares)totheGalacticCepheidsadoptedfrom errorsoftheindividualphotometricpointsareafewmmag. Tammannetal.(2003)(crosses). The first reliable light curve based on photoelectric UBV different telescopes as described in Section 5.1. The ampli- observations was published by Wachmann (1976). Fur- tudes and amplitude ratios are consistent with V1154Cyg ther multicolour photoelectric and CCD photometric data being a Cepheid, e.g. the ratio of the I and V amplitude were published by Szabados (1977), Arellano Ferro et al. is 0.6, which is exactly what is expected.Fig. 5 shows that (1998),Ignatova & Vozyakova(2000),Berdnikov(2008)and theaverageFourierparametersofV1154Cygfitwellallthe Pigulski et al.(2009).Thislatterpapercontainsthedataof progressions, although in each case they are slightly lower a dedicated photometric survey of the whole Kepler field. than themain progression. Space photometric data of V1154Cyg are also available Asacheck,thecoloursofV1154Cygwerecomparedto from the Hipparcos satellite ESA (1997) and the Scientific other Galactic Cepheids in Fig.11. The observed colours of ArchiveoftheOpticalMonitoringCamera(OMC)onboard (B−V)=0.865and(V −I)=1.021werederivedfrom the INTEGRAL.Noneofthesepreviousdatacancompetewith BVRcIc light curves. To correct for extinction, we adopted Kepler in photometric quality. thecolourexcessfromFernie et al.(1995)12 (afterremoving Inaddition,alargenumberofradialvelocitydatahave thesystematictrendinFerniesystemusingtheprescription been collected on V1154Cyg by the Moscow CORAVEL given in Tammann et al. 2003). The extinction-free colours team (Gorynyaet al. 1998). These data obtained between are (B−V)0 = 0.546 and (V −I)0 = 0.612. As shown in 1990−1996 show a slight change in the mean velocity av- Fig.11, these colours fit well within the period-colour rela- eraged over the pulsation cycle, thus Gorynyaet al. (1996) tions definedby theGalactic Cepheids. suspected spectroscopic binarity of this Cepheid. However, radialvelocitydataobtainedbyImbert(1999)muchearlier 6.3 Analysis of the Kepler light curves than the Moscow data and covering a reasonably long time intervaldo not indicate binarity. Withalmostayearofcontinuousdata,itisinprinciplepos- Molenda-Z˙akowicz et al. (2008) determined basic pa- sible to study the stability of the light curve of a classical rameters for our target: [Fe/H] = 0.06 ± 0.07, spectral Cepheid over several dozen pulsation cycles. However, be- type: G2Ib, Teff = 5370±118K, logg = 1.49±0.34, and cause instrumental effects are still present in the data, it is vsini=12.3±1.6kms−1. These parameters are all consis- tooearlytoperformsuchananalysisatleastuntilpixel-level tent with a Cepheid, and place V1154Cyg inside the the- databecomeavailable,whichwouldallowthedatareduction oretical instability strip presented in Fig.1. The chemical to beoptimized for this particular typeof star. composition of V1154Cyg was determined independently The frequency content of the light curve of V1154Cyg by Luck et al. (2006) in a major project of Cepheid spec- was investigated with standard Fourier transform methods troscopy. They published a value of [Fe/H]=−0.10. byapplyingwell-testedsoftwarepackages:SigSpec(Reegen 2007), Period04 (Lenz & Breger 2005) and MuFrAn (Koll´ath 1990). The distorted parts of the LC light curves 6.2 Multicolour photometry of V1154Cyg Fig.10 shows a multicolour light curve for V1154Cyg. It 12 http://www.astro.utoronto.ca/DDO/research/cepheids/ contains 114−120 points in each filter obtained with four /table colourexcess.html 10 R. Szab´o, L. Szabados, C.C. Ngeow, et al. Table 4. Frequencies, amplitudes and phases of the identified frequencypeaksinthefrequencyspectrumofV1154Cygbasedon Q1−Q4LCdata.Theformaluncertainty(1σ)oftheamplitudes isuniformly0.000024. ID freq. σf ampl. ϕphase σϕ d−1 d−1 mag rad rad f0 0.2030244 0.0000002 0.144984 0.794350 0.000027 2f0 0.4060514 0.0000010 0.039291 4.403636 0.000097 3f0 0.6090704 0.0000041 0.009880 1.926416 0.000385 4f0 0.8120798 0.0000375 0.001090 5.812189 0.003496 5f0 1.0151798 0.0000668 0.000611 5.508786 0.006232 6f0 1.2181468 0.0000683 0.000598 3.132366 0.006361 7f0 1.4211946 0.0001077 0.000379 0.442772 0.010040 8f0 1.6241924 0.0001820 0.000224 4.130455 0.016973 9f0 1.8272617 0.0003652 0.000112 1.421814 0.034039 10f0 2.0304263 0.0006842 0.000060 4.893997 0.063769 Figure 12.FrequencyspectrumofV1154CygbasedonQ1−Q4 11f0 2.2334364 0.0001591 0.000026 2.184016 0.148302 LCdata.a)Mainpulsationalfrequencyandthefirsttwoharmon- ics. b) Prewhitened by the three frequency peaks, higher order harmonicsemergeupto10f0.Theremainingpartofthehigher- frequencyrangeofthespectrumwasdividedintotwoforclarity: panel c) shows the frequency range [2.4-15] d−1, and panel d) comprisestherange[15-700]d−1 basedonQ5SCdata.Notethe changeofthescaleontheverticalaxis. have been omitted, because their presence causes spurious frequency peaks around the main frequencies. The affected partsaretheentireQ0andBJD=[2455054.0−2454094.24] 6.4 Behaviour of the pulsation period in Q2. The frequency spectrum shows the main pulsation This is the first occasion that cycle-to-cycle changes in the frequencyatf0 =0.203d−1 andmanyharmonics.Twohar- pulsation period of a Cepheid can be followed. The O − monics(2f0and3f0)areclearlyvisibleintheupperpanelof C analysis of Kepler data of V1154Cyg is published in a Fig. 12.Prewhitening with thesepeaksreveals furtherhar- separatepaper(Derekasetal.2011,inprep.).Herewestudy monicsuptothe10th orderwith verylow amplitudes.This thelong-term behaviourof thepulsation period. is the first time that such high-order harmonics have been The computed times of maxima were calculated from detected, underlining the accuracy of the Kepler observa- theperiodfittedtotheKeplerdata.OneO−Cpointwasde- tions.Theeffectofinstrumentalartefacts(trends,amplitude rivedforthemid-epochofannualsectionsforeachavailable variation) is clearly seen in theremainingpower aroundf0. photometric time series taken from the literature. Besides Apartfromthatthefrequencyspectrumiscompletelyfreeof theKeplermaximaandavailableCCDandphotoelectricob- additional power at thesignificance level up tothe Nyquist servationwepublishforthefirsttimeV1154Cygdatafrom frequency as shown in thelower two panels of Fig.12. digitizedHarvardplatesandeyeestimationsfromSternberg Before finishing this paper, Q5 SC data became avail- AstronomicalInstitute(SAI)photographicplates.Thepass- ableforV1154Cyg.Toinvestigatethehigh-frequencyrange band of these observations is close to Johnson B. We also we used this 94.7d long data set and the 33.5d long SC used visual and CCD observations from the AAVSO Inter- datatakeninQ1.Thetwodatasetshaveaverysimilarfre- nationalDatabase.WherebothJohnsonBandV datawere quency content, and we chose to plot the Q5 SC frequency availableatthesameepoch,weretainedonlytheV maxima. spectrum in panel d of Fig.12,because thelonger timebase These data points are listed in the first column of Ta- ensures betterfrequency resolution and higher SNR. ble5.Column2liststheepochnumber(E =0wasarbitrar- The top of the grass of the remaining spectrum is ilytakenatthefirstmaximumoftheKeplerdata).Weights 5µmag, while the average is 2µmag in the spectrum up were assigned to individual data sets similarly to the case to 50d−1. Above that the top of the grass of the remain- ofV2279Cyg.TheweightsarelistedinColumn4.Thefinal ingspectrumdecreasesto1.5µmag,andfrom100d−1 itre- ephemeriswas derivedbyaweightedlinear least squaresfit mainsconstant.Theaverageoftheremainingpeaksisbelow tothepreliminaryresidualscomputedfromaformerlypub- 1µmaginthishigh-frequencyrange.Theresidualspectrum lished (usually slightly incorrect) period value which is an shows no signal of any shorter period nonradial pulsation inherentstepintheO−C method.Noweightwasassigned modes or solar-like oscillations. tothe photographic normal maxima, theseO−C residuals Thefrequencies,amplitudesandphasesofthedetected wereomittedfromfittingprocedure.TheO−C residualsin andidentifiedpeaksbased on Q1−Q4LCdataarelisted in Column3 havebeen calculated with thefinal ephemeris Table4. The zero epoch was chosen close to moment of the C =BJD(2454955.7260±0.0008)+(4.925454±0.000001)E. firstdatapoint,i.e.BJD=2454954.0. Theerrorshavebeen This period is considered tobe more accurate than theone estimated from Period04. Searching for only one frequency obtained by fitting the frequency and its harmonics. The of the highest amplitude at a time and prewhitening for it source of data is given in the last column of Table5. The andthenrepeatingtheproceduregavepractically thesame O−C diagram isshown in Fig.13. Thisplot indicatesthat results as searching for all theharmonics simultaneously. thepulsation period hasbeen constant duringthelast 40y.

ebook img

Cepheid investigations using the Kepler space telescope PDF

2.4 MB·English
by  R. Szabó
#additional_collections #journals #arxiv
Save to my drive
Quick download
Download
Upgrade Premium
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Cepheid investigations using the Kepler space telescope

Mon.Not.R.Astron.Soc.000,000–000 (0000) Printed13January2011 (MNLATEXstylefilev2.2) Cepheid investigations using the Kepler space telescope R. Szabo´1⋆, L. Szabados1, C.-C. Ngeow2, R. Smolec3, A. Derekas1,4†, 1 P. Moskalik5, J. Nuspl1, H. Lehmann6, G. Fu˝r´esz7, J. Molenda-Z˙akowicz8, 1 0 S. T. Bryson9, A. A. Henden10, D. W. Kurtz11, D. Stello12, J. M. Nemec13, 2 J. M. Benko˝1, L. Berdnikov14,15, H. Bruntt16, N. R. Evans17, N. A. Gorynya18, n a E. N. Pastukhova18, R. J. Simcoe14, J. E. Grindlay19, E. J. Los19, A. Doane19, J S. G. Laycock19, D. J. Mink17, G. Champine19, A. Sliski19, G. Handler3, 2 1 L. L. Kiss1, Z. Kolla´th1, J. Kova´cs20, J. Christensen-Dalsgaard21, H. Kjeldsen21, ] C. Allen22, S. E. Thompson23, J. Van Cleve23 R 1Konkoly Observatory of the Hungarian Academy of Sciences, Konkoly Thege Miklo´s u´t 15-17, H-1121 Budapest, Hungary S 2Graduate Institute of Astronomy, National Central University,Jhongli City, Taoyuan County 32001, Taiwan . 3Institut fu¨r Astronomie, Universita¨t Wien, Tu¨rkenschanzstrasse 17, A-1180 Wien, Austria h 4Department of Astronomy, E¨otv¨os University, Budapest, Hungary p 5Copernicus Astronomical Center, ul. Bartycka18, 00-716, Warsaw, Poland - o 6Thu¨ringer Landessternwarte Tautenburg, Karl-Schwarzschild-Observatorium, 07778 Tautenburg, Germany r 7Harvard-Smithsonian Centerfor Astrophysics, MS 20, 60 Garden Street, Cambridge, MA 02138, USA st 8Astronomical Institute, Universityof Wroc law, ul. Kopernika 11, 51-622, Wroc law, Poland a 9NASA Ames Research Center, Moffett Field, CA 94035, USA [ 10American Association of Variable Star Observers,Cambridge, MA 02138, USA 11Jeremiah Horrocks Institute of Astrophysics, Universityof Central Lancashire, Preston PR1 2HE, UK 1 12Sydney Institute for Astronomy, School of Physics, The University of Sydney, Sydney, 2006 NSW, Australia v 13Department of Physics & Astronomy, Camosun College, Victoria, British Columbia, V8P 5J2, Canada 3 14Sternberg Astronomical Institute, Moscow University,Moscow, Russia 4 15Isaac Newton Institute of Chile, Moscow Branch, UniversitetskijPr. 13, Moscow 119992, Russia 4 16LESIA, UMR 8109, Observatoire de Paris, 92195 Meudon Cedex, France 2 17Smithsonian Astrophysical Observatory, MS4, 60 Garden Street, Cambridge, MA 02138, USA 1. 18Institute of Astronomy, Russian Academy of Sciences, 48 Pyatnitskaya Street, Moscow, 109017, Russia 0 19Harvard College Observatory, 60 Garden Street, Cambridge, MA 02138, USA 1 20Gothard Astrophysical Observatory, H-9707, Szombathely, Hungary 1 21Department of Physics and Astronomy, Aarhus University,DK-8000 Aarhus C, Denmark : 22Orbital SciencesCorporation/NASA Ames Research Center, Moffett Field, CA 94035 v 23SETI Institute/NASA Ames Research Center, MoffettField, CA 94035 i X r a Accepted 2011January12,Received2010December 31 ABSTRACT We report results of initial work done on selected candidate Cepheids to be observed with the Kepler space telescope. Priorto the launch40 candidates were selected from previous surveys and databases. The analysis of the first 322 days of Kepler pho- tometry,andrecentground-basedfollow-upmulticolourphotometryandspectroscopy allowed us to confirm that one of these stars, V1154Cyg (KIC7548061), is indeed a 4.9-dCepheid.Usingthe phaselagmethodweshowthatthisstarpulsatesinthe fun- damentalmode. New radialvelocity data are consistentwith previousmeasurements, suggestingthatalong-periodbinarycomponentisunlikely.No evidenceis seeninthe ultra-precise, nearly uninterrupted Kepler photometry for nonradial or stochastically excited modes at the micromagnitude level. The other candidates are not Cepheids but an interesting mix of possible spotted stars, eclipsing systems and flare stars. Key words: stars: oscillations – stars: variables: Cepheids 2 R. Szab´o, L. Szabados, C.C. Ngeow, et al. 1 INTRODUCTION 0 The Kepler1 space mission is designed to detect Earth- 1 like planets around solar-type stars with the transit method (Borucki et al. 2010) by monitoring continuously 2 over 150000 stars with an unprecedented photometric pre- V1154 Cyg cision. The lifetime of at least 3.5 years and the quasi- g g 3 continuous observations make Kepler an ideal tool to mea- o l sure stellar photometric variability with a precision that 4 is unachievable from the ground (see, e.g. Gilliland et al. 5 2010a). Ground-based photometric observations of Cepheids 6 usually consist of a few (typically one or two) observa- tions per night. Until now it has not been possible to ad- 7 14000 12000 10000 8000 6000 4000 2000 equately cover many consecutive pulsational cycles with Teff ground-based photometry. Ground-based studies of several Figure1.SelectionofCepheidcandidatesbasedonKeplerInput types of variable stars – such as δSct stars, DOV, DBV Catalog effective temperature and logg values (small squares). and DAV white dwarfs, sdBV stars and roAp stars – have Theplotcontains5987kasctargets(redplussigns);theselected benefited from multisite photometric observations. But be- Cepheidcandidatesareshownbyfull(blue)squares.V1154Cyg, cause of their longer period, Cepheids have not been the theonlyconfirmedCepheidsisdenotedbyalarge(blue)square. targetsofsuchconcertedefforts,withthenotableexception ThelinearCepheidinstabilitystripisdenoted bydashedlines. of V473Lyr(Burki et al. 1986). Space-based Cepheid observations were conducted ear- lier with thestar trackeroftheWIREsatellite andtheSo- candidates for the survey phase. First, all known, pos- lar Mass Ejection Imager (SMEI) instrument on board the sible or suspected variable stars in the relevant pul- Coriolis satellite (Bruntt et al. 2008; Spreckley & Stevens sational period range were selected from all available 2008; Berdnikov 2010). The lengths of these data sets (∼ databases containing variability information, such as 1000−1600d) are comparable to the nominal lifetime of GCVS (Samuset al. 2002), ASAS North (Pigulski et al. the Kepler Mission. The WIRE and SMEI data of Polaris 2009), ROTSE (Akerlof et al. 2000) and HAT catalogs (αUMi,V =2.005) confirmedthatthepulsationamplitude (Hartman et al. 2004). Light curve shapes and the logP of Polaris has been increasing again, after a long period of vs. J −H diagram (Pojmanski & Maciejewski 2004) were decrease. Kepler is capable of delivering comparably accu- also utilized for further selection. Finally, stars with close ratephotometricobservationsforthemuchfainterCepheid, (bright)companionswereexcluded.Thisprocedureresulted V1154Cyg (V =9.19). inalistof26stars.Wenotethatthesecatalogsarenotcom- The Kepler Asteroseismic Science Consortium kasc2 pletein coverage of thefield,norin therelevant magnitude wassetuptoexploittheKeplerdataforstudyingstellarpul- range. sations. kasc working group number 7 (WG7) is dedicated Asasecondapproach,wesearchedforstarslyinginside to the investigation of Cepheids. In compliance with one of orclosetotheCepheidinstabilitystripbasedontheKepler the original kasc proposals submitted before the launch of Input Catalog (KIC)3 effective temperature and logg val- the space telescope, we searched for Cepheids among a list ues.StarsfainterthanKp=16.0magwereexcluded,where of 40 targets, including the only previously known Cepheid KpdenotestheKeplermagnitudesystem(seeSection3.for V1154Cyg (KIC7548061) in the field. In this paper we de- moredetails on theKepler magnitudes). Candidateswith a scribethefirstresultsfromKeplerobservationsofaCepheid contaminationindex(CI)largerthan0.5werealsoomitted, and a couple of not confirmed Cepheids, complemented by whereCI=0 meansthat all thefluxin theaperturecomes extendedground-based follow-up observations. fromthetargetandincaseofCI=1thefluxentirelycomes fromsurroundingsources.Thisresultedin14additionaltar- getsasshowninFig.1.ThelinearCepheidinstabilitystrip, denoted by dashed lines in the figure, was calculated with 2 TARGET SELECTION theFlorida-Budapest code (Szab´o et al. 2007). We note that the process followed to derive the KIC Being supergiants, Cepheids are rare and the number of CepheidsintheKeplerfixed105squaredegreefieldofview parameters was optimized to find main-sequencestars with high probability and to distinguish them from cool giants (FOV) is expected to be low. Due to telemetry bandwidth constraints,Keplercannotobserveallstarsinthefield.The (Batalha et al. 2010). The precision of the parameters is therefore limited for our purposes. This is illustrated in first 10 months of operation (observing quarters Q0–Q4) Fig.1 by the fact the KIC parameters of V1154Cyg (the were therefore dedicated to a survey phase to find the best only previously known Cepheid in the field) put it outside candidates,whichwouldthenbeobservedduringtherestof thecomputed instability strip by several hundredKelvins. themission. Consequently, we proposed to observe the above men- We used two different approaches to find Cepheid tioned 40 stars by Kepler during the‘survey period’. 1 http://kepler.nasa.gov 2 http://astro.phys.au.dk/KASC/ 3 http://archive.stsci.edu/kepler Cepheid investigations using the Kepler space telescope 3 Table1.MainpropertiesoftheobservedKeplerCepheidandCepheidcandidates.Theuncertaintyoftheperiodisgiveninparentheses andreferstothelastdigit(s)oftheperiod.Q3.1meansthefirstmonthofQ3.Seetextforthedefinitionofthecontamination index. KICID Othername α δ Kp Period Contam. Runs J2000 J2000 mag days Index 7548061 V1154 Cyg 19 48 15.45 +43 07 36.77 8.771 4.925454(1) 0.036 LC: Q01234 SC: Q1 2968811 ASAS190148+3807.0 190148.21 +380701.99 13.469 14.8854(12) 0.047 LC:Q01234 6437385 ASAS192000+4149.1 192000.12 +414907.46 11.539 13.6047(7) 0.075 LC:Q01234 8022670 V2279Cyg 191854.46 +434925.82 12.471 4.12564(5) 0.189 LC:Q1234 12406908 ROTSE1 192344.99 +511611.75 12.354 13.3503(45) 0.015 LC:Q01SC:Q3.1 J192344.95+511611.8 tions are based on the same 6-s integrations which are summedtoformtheLCandSCdataonboard.Inthiswork we use BJD-corrected, raw LC data (Jenkins et al. 2010) spanningfromQ0toQ4,i.e.321.7dofquasi-continuousob- servations. Some of our targets (V1154Cyg among them) wereobservedinSCmodeaswell.Weexploitthisopportu- nity to compare LC and SC characteristics and investigate thefrequencyspectrumtoamuchhigherNyquistfrequency (733.4d−1 vs. 24.5d−1). The saturation limit is between Kp ≃ 11 − 12mag depending on the particular chip the star is observed; brighter than this, accurate photometry can be performed up to Kp ≃ 7mag with judiciously designed apertures (Szab´oet al. 2010). SinceV1154Cyg is much brighter than the saturation limit, it required special treatment, as can be seen in Fig.2, which shows a 50 pixel box centered on V1154Cyg. The plot was made using keplerffi4 written byM. Still. To illustrate some of the common characteristics of the data we show in Fig.3 the raw Kepler light curve of V1154Cyg after normalizing the raw flux counts and con- Figure2.VicinityofV1154CygfromoneoftheFullFrameIm- verting the fluxes to the magnitude scale. The small gaps ages.Thestarisslightlysaturatedasindicatedbyitsbrightness in thelight curvearedueto unplannedsafe modeandloss- andtheelongatedshape. of-fine-point events, as well as regular data downlink peri- ods.ThisLCdatasetspanningQ0–Q4datacontains14485 points. Fig. 4 shows a 33.5d segment (Q1) where SC data 3 KEPLER OBSERVATIONS are available for V1154Cyg containing 49032 data points. ThevaryingamplitudeseeninFig. 3isofinstrumental Kepler was launched on 2009 March 6 into a 372-d solar origin. It is a result of the small drift of the telescope, cou- orbit,andisobservinga105squaredegreeareaoftheskyin pled with different pixel sensitivities. In addition, different betweentheconstellationsofCygnusandLyra(Koch et al. aperture masks are assigned to the targets quarterly which 2010). After a 9.7d commissioning phase (Q0), the regular result in small changes in the measured flux. The most no- observations started on 2009 May 12. In order to ensure tableamplitudechangeisseentowardstheendofQ2,which optimal solar illumination of the solar panels, a 90degree was noted to be due to flux flowing outside of the optimal rollof thetelescope isperformed at theendofeach quarter aperture(bleeding),affectingthemeasuredbrightness.For- of its solar orbit. The first quarter lasted only for 33.5d tunately, in Q2 a larger mask was also downloaded besides (Q1), while subsequent quarters are all three months long. the standard optimal aperture assigned to this star which Ineach of thefourquartersannuallytheKeplertargets fall allowed us to investigate the variation of flux outside the on different CCDs. optimalaperture.Thisconfirmedthatthetotalfluxwasin- The Kepler magnitude system (Kp) refers to the wide deedcapturedwithinthelargeraperture,andhencethatthe passband (430−900nm) transmission of the telescope and star shows no intrinsic amplitude variation within the cur- detector system. Note that the Kepler magnitudes (Kp) rent accuracy of the data. Without further information we werederivedbeforethemissionlaunchandareonlyapprox- cannotchoosebetweenapossibleslightamplitudevariation imate values. Currently Kepler processing does not provide and instrumental effects in other quarters. calibrated Kepler magnitudes Kp (Kolenberg et al. 2010). Both long cadence (LC, 29.4min, Jenkins et al. 2010), and short cadence (SC, 58.9s, Gilliland et al. 2010b) observa- 4 http://keplergo.arc.nasa.gov/ContributedSoftwareKeplerFFI.shtml 4 R. Szab´o, L. Szabados, C.C. Ngeow, et al. 0.25 0.20 Q0 Q1 Q2 Q3 Q4 0.20 0.15 0.15 0.10 0.10 0.05 0.05 Kp Kp 0.00 ∆ 0.00 ∆ -0.05 -0.05 -0.10 -0.10 -0.15 -0.15 -0.20 -0.20 54950 55000 55050 55100 55150 55200 55250 54965 54970 54975 54980 54985 54990 54995 55000 BJD-2,400,000 BJD-2,400,000 Figure 3. Raw Kepler light curve of V1154Cyg. The varying Figure 4. Q1 SC Kepler light curve of V1154Cyg. It is almost amplitudeisinstrumentalinorigin;seetextformoredetails. indistinguishable from LC data taken during the same time in- terval.Notetheessentiallyuninterruptedsampling. 5.1 Multicolour photometry 4 TARGET CLASSIFICATION Ground-based multicolour observations provide additional After the release of the first Kepler light curves to the information, therefore complement the space-born photom- kasc it turned out that only a few stars showed Cepheid etrytakenin ‘whitelight’.Weusedthefollowing telescopes likevariability.ThepropertiesoftheseKeplerCepheidcan- to gather BVR I Johnson-Cousins magnitudes: c c didates are listed in Table1. Periods were determined for the Kepler LC light curves using the period04 program (i) Lulin One-meter Telescope (LOT) at Lulin Obser- (Lenz& Breger 2005). vatory (Taiwan)5: 1-m Cassegrain-telescope with PI1300B Apart from these candidates a number of other vari- CCD; able stars were included in the initial sample of 40 stars. (ii) SLT at Lulin Observatory (Taiwan): 0.4-m RC-tele- Based on their light curvesand theautomatic classification scope with Apogee U9000 CCD; (Blomme et al. 2010) we classify these as eclipsing binaries (iii) Tenagra telescope at Tenagra II Observatory6 (8 stars), ellipsoidal variables (6), delta Scuti stars (3), an (USA):robotic 0.8-m RC-telescope with SITe CCD; SPB star (1), long period variables (7) and stars with no (iv) SonoitaResearchObservatory(SRO,USA)7,0.35-m obviousvariations (7).While this sample is arich source of robotic telescope with a SBIG STL-1001E CCD. flaring, spotted, granulated stars, thedetailed investigation Observations were performed from September 2009 to ofthesenon-Cepheidsisoutofscopeofthispaper.Inaddi- August2010.Alloftheimagingdatawerereducedinastan- tion, we inspected all 6300 kasc light curves, but found no dardway(bias-subtracted,dark-subtractedandflat-fielded) Cepheids among them. using iraf8. Instrumental magnitudes for the stars in the Three originally proposed stars were not observed due imagesweremeasuredusingSExtractor(Bertin & Arnouts to technical reasons (position on the CCD chip, bright- 1996) with aperture photometry. For the SRO data a sep- ness/faintness, contamination orpixelnumberconstraints). arate photometric pipeline based on daophot was used, The large variety of non-Cepheids in our sample indi- whichincludedanalysisofanensembleofcomparison stars. cates that the spatial resolution and photometric accuracy Theinstrumentalmagnitudesweretransformedtothestan- of ASAS was not adequate to select such relatively faint dard magnitudes using Landolt standards (Landolt 2009). Cepheids.ItfurthershowsthattheKeplerInputCatalog is Times of observation were converted to heliocentric Julian clearly not optimal for our purpose. day. We gathered between 80 and 120 frames per target for eachpassband,excludingKIC12406908whichwasobserved only at TNG and hence had fewer images taken. We have found a small offset between SRO and other B data in the 5 GROUND-BASED FOLLOW UP case of V1154Cyg. Fortunately, data were taken in TNG and SRO very close in time (less than 1 minute) twice, al- To further confirm or disprove the Cepheid nature of our lowing us to correct the SRO data by the shift measured candidatesinTable1weemployedground-basedmulticolour photometry and in some cases spectroscopy. In the case of 5 http://www.lulin.ncu.edu.tw/english/index.htm V1154Cyg regular radial velocity observations and multi- 6 http://www.tenagraobservatories.com colour photometry were scheduled, which helped us to gain 7 http://www.sonoitaobservatories.org more information on the pulsation. In the following we de- 8 irafisdistributedbytheNational Optical AstronomyObser- scribe these ground-based follow-up observations and the vatories,whichareoperatedbytheAssociationofUniversitiesfor subsequent scrutinizing classification process on a star-by- Research in Astronomy, Inc., under cooperative agreement with star basis. theNationalScienceFoundation. Cepheid investigations using the Kepler space telescope 5 0.55 7.5 0.50 7.0 0.45 6.5 0.40 6.0 0.35 5.5 0.30 5.0 R21 0.25 φ21 4.5 0.20 4.0 0.15 KIC 2968811 3.5 KIC 6437385 0.10 KIC 12406908 3.0 0.05 V2279 Cyg 2.5 V1154 Cyg 0.00 2.0 0 5 10 15 20 25 30 0 5 10 15 20 25 30 P [days] P [days] 0.40 14.0 0.35 12.0 0.30 10.0 0.25 8.0 R31 0.20 φ31 6.0 0.15 4.0 0.10 0.05 2.0 0.00 0.0 0 5 10 15 20 25 30 0 5 10 15 20 25 30 P [days] P [days] Figure5.Fundamentalmode(redfilled)andfirstovertone(blueopensymbols)FourierparameterprogressionsinJohnsonV passband forGalacticCepheidsample.V1154CygandotherKeplerCepheidcandidatesarealsoshownbasedontheirnewground-basedJohnson V photometry. Error bars for V1154Cyg and V2279Cyg are comparable to, or smaller than the symbol size, so these were omitted from the plot.The Galactic Cepheid light curve parameters were compiled from Antonello&Lee (1981), Moffett&Barnes (1985), Antonello&Poretti (1986), Antonello,Poretti&Reduzzi (1990), Mantegazza &Poretti (1992), Poretti (1994), Antonello&Morelli (1996). Cepheid. We discuss our findings for individual candidates Table 2. Ground-based multicolour photometry of V1154Cyg andKeplerCepheidcandidates.Theentiretableisavailableonly below. electronically. KICID HJD Mag. Err. Filt. Obs. 5.2 Spectroscopy 02968811 2455103.7700 14.857 0.013 B SRO Spectroscopic observations were conducted to measure ra- 02968811 2455103.7809 14.823 0.013 B SRO 02968811 2455104.7544 14.797 0.011 B SRO dial velocity and derive stellar parameters. We obtained ... ... ... ... ... ... spectra of V1154Cyg with the Coude-Echelle spectrograph attachedtothe2-mtelescopeoftheThu¨ringerLandesstern- warte (TLS) Tautenburg. Spectra cover the region 470 − 740nm, the spectral resolution is R=33000, the exposure in these epochs: BSRO−BTNG =0.054. All the (corrected) time was 30min per spectrum. Spectra were reduced using photometric measurements are available online, while Ta- standard midas packages. The reduction included the re- ble2 shows the layout of thedata. moval of cosmic rays, bias and stray light subtraction, flat We decomposed the multicolour light curves to derive fielding, optimum order extraction, wavelength calibration the widely used Fourier parameters Ri1 and φi1, as de- using a Th-Ar lamp, normalization to the local continuum, fined by Simon & Lee (1981), which characterize the light and merging of the orders. Small instrumental shifts were curve shape. The Johnson V results are plotted in Fig.5 correctedbyanadditionalcalibrationusingalargernumber alongwiththeFourierparametersoffundamentalmode(red oftelluricO2 lines.Inaddition,onespectrumofV1154Cyg filled points) and first overtone (blue open circles) Galactic was taken on 2007 June 15 at the M. G. Fracastoro sta- Cepheids as a function of pulsational period. As all Galac- tion (SerraLaNave,Mt. Etna) oftheINAF - Osservatorio tic Cepheids follow the main Fourier progressions in Fig.5, Astrofisico di Catania (INAF-OACt). We used the 91-cm a star that lies outside the overall pattern is unlikely to be telescope and fresco, the fiber-fed reosc echelle spectro- 6 R. Szab´o, L. Szabados, C.C. Ngeow, et al. 0.10 0.05 KIC 2968811 0.08 0.04 KIC 6437385 0.06 0.03 0.04 0.02 0.02 0.01 Kp 0.00 Kp 0.00 ∆ ∆ -0.02 -0.01 -0.04 -0.02 -0.06 -0.03 -0.08 -0.04 -0.10 -0.05 55235 55240 55245 55250 55255 55260 55265 55270 55275 55235 55240 55245 55250 55255 55260 55265 55270 55275 BJD-2,400,000 BJD-2,400,000 0.30 0.15 KIC 8022670 V2279 Cyg KIC 12406908 0.20 0.10 0.10 0.05 Kp 0.00 Kp 0.00 ∆ ∆ -0.10 -0.05 -0.20 -0.10 -0.30 -0.15 55024 55026 55028 55030 55032 55034 55036 55090 55095 55100 55105 55110 55115 55120 55125 BJD-2,400,000 BJD-2,400,000 Figure 6.Representative partsof Keplerphotometry offour Cepheidcandidates (excluding theknown CepheidV1154Cyg). LC light curvesareplotted,exceptforKIC12406908whereonemonthofSCdataisplotted.The‘shoulders’onthedescendingbranchesindicate spottedstarsandtheflaresaretypicalforactivestars.TheselightcurvesindicatethatKIC8022670(V2279Cyg)isthemostpromising Cepheidcandidate ofthefour. graph which allowed us to obtain spectra in the range of first panel of Fig.7. We note that in the case of a Cepheid 4300−6800˚A with a resolution R=21000. in this period range the bump should be present on the For V2279Cyg we made a high-resolution spectrum ascending branch. By carefully examining the short bright- in a single-shot 1680 s exposure with the 1.5-m telescope eningseenintheKeplerlightcurveweexcludeinstrumental at the Fred Lawrence Whipple Observatory (Mt. Hopkins, or other external origin based on known artifacts discussed Arizona) using the Tillinghast Reflection Echelle Spectro- intheKeplerDataReleaseNotes9.Takingintoaccountthe graph(TRES)on2010September24.Thespectrumiscross- strong flare events (Fig.6), we conclude that this object is dispersed withawavelength rangeof3860−9100˚Aover51 definitely not a Cepheid. spectral orders. KIC6437385 = ASAS 192000+4149.1 (P = 13.6047d). This star shows a similar light curve to KIC2968811 with an amplitude of 0.09mag (Fig.6 upper 5.3 Remarks on individual candidates right panel). Within the uncertainty the Fourier parame- In the following we summarize the Kepler light curve char- ters R21 matches the progressions, but R31 does not. The phases are close to the phases of the bulk of the Galactic acteristics, the multicolour photometry, the Fourier param- Cepheids(Fig.5).TheB amplitudeissmallerthantheam- eters and the spectroscopic properties of the four Cepheid plitudemeasuredinV asshowninFig.7,thisisinconsistent candidates (see Table1), before moving to the previously with the Cepheid classification. The light variation shows known Cepheid (V1154Cyg) in Sect. 6. thesame shoulders on thedescendingbranch asin thecase KIC2968811 = ASAS 190148+3807.0 (P = ofKIC2968811,whileitsamplitudeandshapearechanging 14.8854d). In principle the light curve shape matches all throughout the more than 300d long Kepler observations. the Fourier parameter progressions within the uncertainty, In addition, several flares can be seen that are intrinsic to although in the case of φ21 and R31 it is a border line thestar.Oneisshown inFig.6.Thus,KIC6437385 ismost case (Fig.5).Theamplitudeis0.19mag intheKeplerpass- probably not a Cepheid. band.‘Shoulders’orbumpsappearonthedescendingbranch (Fig.6,upperleftpanel),buttheyarenotpresentatthebe- ginningoftheobservations.This effectis frequentlyseenin spotted stars. The same effect may cause the larger scatter of the folded ground-based multicolour light curves in the 9 http://archive.stsci.edu/kepler/data release.html Cepheid investigations using the Kepler space telescope 7 Figure 7. Phased BVRcIc Johnson-Cousins phase-folded light curves of three Cepheid candidates taken during the 2010 observing season: KIC2968811 (left panel), KIC6437385 (middle panel) and KIC8022670 (right panel). For clarity, error bars were plotted only fordatapointswitherrorslargerthan0.02mag. KIC12406908=ROTSE1J192344.95+511611.8 thephotometricsurvey byPigulski et al. (2009) resulted in (P = 13.3503d). Interestingly, this variable star exhibits usefuldataforfollowingtheperiodchangesofthisstar.Ad- verysimilarlightcurvecharacteristicstothepreviouslydis- ditionalphotometricdataarealsoavailablefromtheSuper- cussed objects (Fig.6, bottom right panel). It has an am- WASPpublicarchive10andtheScientificArchiveoftheOp- plitudeof0.3magin theKeplerpassband.Wehavetoofew tical Monitoring Camera (OMC) on board INTEGRAL11. data points from ground-based observations to put strong Both these archives contain data obtained in a single pho- constraintsontheFourierparameters,butdespitethelarge tometric band, and we used these data to investigate the uncertainties this object seems to be slightly off the main period behaviourof V2279Cyg. location expected for Cepheids (Fig.5). The ratio of I and TheKeplerlightcurveofV2279Cygseemsstable,with- c V amplitudes is 0.74, which is close to, but slightly larger, outanynoticeable light curvechanges (partof theQ2light thanthestandard0.6.TheSClightcurvesshowsmallshort- curve is plotted in Fig.6). We only see long term varia- duration flares at BJD2455106.1 and 2455110.6, and a tions similar in amplitude to what we noted for V1154 Cyg largeoneatBJD2455118.2whichshowedacomplex,multi- (Fig.3), which in this case might also just be instrumen- peakedmaximumlastingfor7h.Wenotethatafewoutliers tal effects. What makes V2279Cyg particularly suspicious were removed from the Kepler light curve, which however is the presence of many flares, one of them is clearly seen does not influence our conclusions. Based on the observed at BJD2455030.0 in Fig.6. The (almost) strictly periodic features we can safely exclude this star from our Cepheid variations and the value of the period is consistent with a sample. rotational modulation. KIC8022670 = V2279Cyg (P =4.12564d). Based Multicolour photometry can be decisive in solving this on the space and ground-based follow-up observations this classification problem because Cepheids have characteristic starisastrongCepheidcandidate.Wethereforelookedinto amplitude ratios. Our new photometric data suggest that the literature for a more rigorous assessment of whether V2279CygisnotaCepheid.Theamplitudeofitsbrightness Cepheid pulsation is the cause of the observed variability. variation in V is only slightly smaller than the amplitude Variability of this star was revealed by Dahlmark (2000) in the B band: the ratio is about 0.9 instead of the usual during a photographic search for variable stars, and in- valueof0.65−0.70 (Klagyivik & Szabados2009).Although dependently by Akerlof et al. (2000) as a result of the abrightbluecompanionstarisabletosuppresstheobserv- ROTSE project. While Akerlof et al. (2000) classified the able amplitude of the light variations in the B band, the star as a Cepheid with a period of 4.12298d, Dahlmark observed amplitude ratio is incompatible with the Cepheid (2000) interpreted it as an RS CVn type variable due to natureevenif thestar had a veryhot companion. R21,R31 its proximity to a known X-ray source detected by the and φ31 place it among the first overtone objects, but φ21 ROSAT mission (Voges et al. 1999). This object, ROTSE is very different from both fundamental and first overtone J191853.61+434930.0 = LD349, was finally designated progressions (Fig.5). as V2279Cyg among the variable stars (Kazarovets et al. In a next step we analyze the cycle-to-cycle variations 2003). in the periodicity using the so-called O−C diagram. The WefirstlookedatROTSE-Iphotometricdataretrieved O−C diagram constructed for the moments of brightness from the NSVS data base (Wo´zniak et al. 2004), but con- maximaisplottedinthetoppanelofFig.8.Thezeroepoch clude that this data set does not allow us to unambigu- was arbitrarily chosen at the first maximum in the Kepler ously classify the star. A dedicated photometric project to dataset. A least squares fit totheO−C residuals resulted select TypeII Cepheids among the ROTSE-I targets was performed by Schmidtet al. (2007). Based on their two- colour(V,R)photometrytheyconcludedthatV2279Cygis 10 http://www.wasp.le.ac.uk/public/index.php aprobableCepheid with aperiod of4.117d.More recently, 11 http://sdc.laeff.inta.es/omc/index.jsp 8 R. Szab´o, L. Szabados, C.C. Ngeow, et al. 200 1000 Na D Li Ca II H 800 150 1000 U U 600 U D 100 D D A A 400 A 500 50 200 0 0 0 3940 3960 3980 4000 5860 5880 5900 5920 5940 6660 6690 6720 6750 wavelength (Angstrom) wavelength (Angstrom) wavelength (Angstrom) Figure 9.ThetresspectrumofV2279Cyg.Fromlefttoright:zoom intoCaIIH(3968.5˚A),NaD(5890and5896˚A),Li(6708˚A). Table 3. Sample table for O−C residuals for V2279 Cyg. The entiretableisavailableonline. JD⊙ − E O−C W Reference 2400000 [d] 51474.9729 −843 −1.7218 2 NSVS 52626.3311 −564 −1.4177 1 Integral OMC 53051.2057 −461 −1.4842 3 Schmidtetal. 53368.6341 −384 −1.7304 1 Schmidtetal. 54265.3816 −191 −1.2317 1 SWASP 54645.2207 −75 0.0030 1 SWASP 54954.6214 0 0.0105 5 Kepler ... ... ... ... ... plotted three segments of the spectrum taken with the tres spectrograph in Fig.9 containing these lines: CaiiH (3968.5˚A),NaD (5890 and 5896˚A) and Li(6708˚A).While the NaD lines are normal, the characteristics of the two other lines do not support a Cepheid nature of V2279Cyg. TheLi(6708˚A)isneverseeninemissioninCepheids,while theCaemissionimplieschromosphericactivity,whichisalso Figure 8. Top panel: O−C diagram of V2279Cyg for the Ke- notaCepheidcharacteristic. Wefittedtheoreticaltemplate plerdata.Bottom panel:O−C diagramofV2279Cyginvolving spectra from the extensive spectral library of Munari et al. all photometric data. The point size corresponds to the weight (2005)tothespectrum anddeterminedthefollowing atmo- assignedtothemaxima. spheric parameters: Teff = 4900±200K, logg = 3.7±0.4, [M/H] = −1.2±0.4 and vsini = 40kms−1. The resulting parameters are also incompatible with a Cepheid variable, in the best fitting period of 4.125642d, which is somewhat but suggest a cool main-sequence star with moderate rota- longerthananyformerlypublishedvalueforV2279Cyg.We tion. thenincludedallground-basedobservationswecouldfindto Summarizing the previous subsections, we conclude construct a newO−C diagram using the ephemeris that all the candidates turned out not to be Cepheids ex- C =BJD(2454954.6109±0.0023)+(4.125642±0.000051)E. cept the already known Cepheid V1154Cyg, which we will Table3 contains the moments of maxima, the correspond- describe in detail in the following. ing epochs and the O−C residuals of the available obser- vations. We assigned different weights to data from differ- ent sources. These weights (between 1 and 5) correspond 6 V1154 Cyg THE ONLY KEPLER CEPHEID to the ‘goodness’ of the seasonal light curve, larger num- ber means better coverage and smaller scatter (for Kepler In this section we analyze both Kepler data and ground- data W=5). The period of V2279Cyg show large fluctua- basedfollow-up observationsof what isapparentlytheonly tions during the last decade (see the lower panel of Fig.8). Cepheid being observed in Kepler’s FOV. Our additional ground-based photometric data indicate a very recent change in the period that we will investigate 6.1 Observational data prior Kepler using future quartersof Kepler data. Wenote that the star hasahighcontaminationindex(Table1),indicatingthatthe Brightness variation of V1154Cyg was discovered by ∼18percentofthemeasuredfluxcomesformothersources. Strohmeier et al. (1963). Cepheid type variation and a pe- ThespectrumofV2279Cygofferstheclearest evidence riodicity somewhat shorter than 5d were obvious from that this star has been misclassified as a Cepheid. We have the photographic magnitudes leading to the discovery. Cepheid investigations using the Kepler space telescope 9 Figure 10. BVRcIc Johnson-Cousins phase-folded light curves Figure 11. Comparison of the extinction-corrected colours of of V1154Cyg taken during the 2010 observing season. Typical V1154Cyg(filledsquares)totheGalacticCepheidsadoptedfrom errorsoftheindividualphotometricpointsareafewmmag. Tammannetal.(2003)(crosses). The first reliable light curve based on photoelectric UBV different telescopes as described in Section 5.1. The ampli- observations was published by Wachmann (1976). Fur- tudes and amplitude ratios are consistent with V1154Cyg ther multicolour photoelectric and CCD photometric data being a Cepheid, e.g. the ratio of the I and V amplitude were published by Szabados (1977), Arellano Ferro et al. is 0.6, which is exactly what is expected.Fig. 5 shows that (1998),Ignatova & Vozyakova(2000),Berdnikov(2008)and theaverageFourierparametersofV1154Cygfitwellallthe Pigulski et al.(2009).Thislatterpapercontainsthedataof progressions, although in each case they are slightly lower a dedicated photometric survey of the whole Kepler field. than themain progression. Space photometric data of V1154Cyg are also available Asacheck,thecoloursofV1154Cygwerecomparedto from the Hipparcos satellite ESA (1997) and the Scientific other Galactic Cepheids in Fig.11. The observed colours of ArchiveoftheOpticalMonitoringCamera(OMC)onboard (B−V)=0.865and(V −I)=1.021werederivedfrom the INTEGRAL.Noneofthesepreviousdatacancompetewith BVRcIc light curves. To correct for extinction, we adopted Kepler in photometric quality. thecolourexcessfromFernie et al.(1995)12 (afterremoving Inaddition,alargenumberofradialvelocitydatahave thesystematictrendinFerniesystemusingtheprescription been collected on V1154Cyg by the Moscow CORAVEL given in Tammann et al. 2003). The extinction-free colours team (Gorynyaet al. 1998). These data obtained between are (B−V)0 = 0.546 and (V −I)0 = 0.612. As shown in 1990−1996 show a slight change in the mean velocity av- Fig.11, these colours fit well within the period-colour rela- eraged over the pulsation cycle, thus Gorynyaet al. (1996) tions definedby theGalactic Cepheids. suspected spectroscopic binarity of this Cepheid. However, radialvelocitydataobtainedbyImbert(1999)muchearlier 6.3 Analysis of the Kepler light curves than the Moscow data and covering a reasonably long time intervaldo not indicate binarity. Withalmostayearofcontinuousdata,itisinprinciplepos- Molenda-Z˙akowicz et al. (2008) determined basic pa- sible to study the stability of the light curve of a classical rameters for our target: [Fe/H] = 0.06 ± 0.07, spectral Cepheid over several dozen pulsation cycles. However, be- type: G2Ib, Teff = 5370±118K, logg = 1.49±0.34, and cause instrumental effects are still present in the data, it is vsini=12.3±1.6kms−1. These parameters are all consis- tooearlytoperformsuchananalysisatleastuntilpixel-level tent with a Cepheid, and place V1154Cyg inside the the- databecomeavailable,whichwouldallowthedatareduction oretical instability strip presented in Fig.1. The chemical to beoptimized for this particular typeof star. composition of V1154Cyg was determined independently The frequency content of the light curve of V1154Cyg by Luck et al. (2006) in a major project of Cepheid spec- was investigated with standard Fourier transform methods troscopy. They published a value of [Fe/H]=−0.10. byapplyingwell-testedsoftwarepackages:SigSpec(Reegen 2007), Period04 (Lenz & Breger 2005) and MuFrAn (Koll´ath 1990). The distorted parts of the LC light curves 6.2 Multicolour photometry of V1154Cyg Fig.10 shows a multicolour light curve for V1154Cyg. It 12 http://www.astro.utoronto.ca/DDO/research/cepheids/ contains 114−120 points in each filter obtained with four /table colourexcess.html 10 R. Szab´o, L. Szabados, C.C. Ngeow, et al. Table 4. Frequencies, amplitudes and phases of the identified frequencypeaksinthefrequencyspectrumofV1154Cygbasedon Q1−Q4LCdata.Theformaluncertainty(1σ)oftheamplitudes isuniformly0.000024. ID freq. σf ampl. ϕphase σϕ d−1 d−1 mag rad rad f0 0.2030244 0.0000002 0.144984 0.794350 0.000027 2f0 0.4060514 0.0000010 0.039291 4.403636 0.000097 3f0 0.6090704 0.0000041 0.009880 1.926416 0.000385 4f0 0.8120798 0.0000375 0.001090 5.812189 0.003496 5f0 1.0151798 0.0000668 0.000611 5.508786 0.006232 6f0 1.2181468 0.0000683 0.000598 3.132366 0.006361 7f0 1.4211946 0.0001077 0.000379 0.442772 0.010040 8f0 1.6241924 0.0001820 0.000224 4.130455 0.016973 9f0 1.8272617 0.0003652 0.000112 1.421814 0.034039 10f0 2.0304263 0.0006842 0.000060 4.893997 0.063769 Figure 12.FrequencyspectrumofV1154CygbasedonQ1−Q4 11f0 2.2334364 0.0001591 0.000026 2.184016 0.148302 LCdata.a)Mainpulsationalfrequencyandthefirsttwoharmon- ics. b) Prewhitened by the three frequency peaks, higher order harmonicsemergeupto10f0.Theremainingpartofthehigher- frequencyrangeofthespectrumwasdividedintotwoforclarity: panel c) shows the frequency range [2.4-15] d−1, and panel d) comprisestherange[15-700]d−1 basedonQ5SCdata.Notethe changeofthescaleontheverticalaxis. have been omitted, because their presence causes spurious frequency peaks around the main frequencies. The affected partsaretheentireQ0andBJD=[2455054.0−2454094.24] 6.4 Behaviour of the pulsation period in Q2. The frequency spectrum shows the main pulsation This is the first occasion that cycle-to-cycle changes in the frequencyatf0 =0.203d−1 andmanyharmonics.Twohar- pulsation period of a Cepheid can be followed. The O − monics(2f0and3f0)areclearlyvisibleintheupperpanelof C analysis of Kepler data of V1154Cyg is published in a Fig. 12.Prewhitening with thesepeaksreveals furtherhar- separatepaper(Derekasetal.2011,inprep.).Herewestudy monicsuptothe10th orderwith verylow amplitudes.This thelong-term behaviourof thepulsation period. is the first time that such high-order harmonics have been The computed times of maxima were calculated from detected, underlining the accuracy of the Kepler observa- theperiodfittedtotheKeplerdata.OneO−Cpointwasde- tions.Theeffectofinstrumentalartefacts(trends,amplitude rivedforthemid-epochofannualsectionsforeachavailable variation) is clearly seen in theremainingpower aroundf0. photometric time series taken from the literature. Besides Apartfromthatthefrequencyspectrumiscompletelyfreeof theKeplermaximaandavailableCCDandphotoelectricob- additional power at thesignificance level up tothe Nyquist servationwepublishforthefirsttimeV1154Cygdatafrom frequency as shown in thelower two panels of Fig.12. digitizedHarvardplatesandeyeestimationsfromSternberg Before finishing this paper, Q5 SC data became avail- AstronomicalInstitute(SAI)photographicplates.Thepass- ableforV1154Cyg.Toinvestigatethehigh-frequencyrange band of these observations is close to Johnson B. We also we used this 94.7d long data set and the 33.5d long SC used visual and CCD observations from the AAVSO Inter- datatakeninQ1.Thetwodatasetshaveaverysimilarfre- nationalDatabase.WherebothJohnsonBandV datawere quency content, and we chose to plot the Q5 SC frequency availableatthesameepoch,weretainedonlytheV maxima. spectrum in panel d of Fig.12,because thelonger timebase These data points are listed in the first column of Ta- ensures betterfrequency resolution and higher SNR. ble5.Column2liststheepochnumber(E =0wasarbitrar- The top of the grass of the remaining spectrum is ilytakenatthefirstmaximumoftheKeplerdata).Weights 5µmag, while the average is 2µmag in the spectrum up were assigned to individual data sets similarly to the case to 50d−1. Above that the top of the grass of the remain- ofV2279Cyg.TheweightsarelistedinColumn4.Thefinal ingspectrumdecreasesto1.5µmag,andfrom100d−1 itre- ephemeriswas derivedbyaweightedlinear least squaresfit mainsconstant.Theaverageoftheremainingpeaksisbelow tothepreliminaryresidualscomputedfromaformerlypub- 1µmaginthishigh-frequencyrange.Theresidualspectrum lished (usually slightly incorrect) period value which is an shows no signal of any shorter period nonradial pulsation inherentstepintheO−C method.Noweightwasassigned modes or solar-like oscillations. tothe photographic normal maxima, theseO−C residuals Thefrequencies,amplitudesandphasesofthedetected wereomittedfromfittingprocedure.TheO−C residualsin andidentifiedpeaksbased on Q1−Q4LCdataarelisted in Column3 havebeen calculated with thefinal ephemeris Table4. The zero epoch was chosen close to moment of the C =BJD(2454955.7260±0.0008)+(4.925454±0.000001)E. firstdatapoint,i.e.BJD=2454954.0. Theerrorshavebeen This period is considered tobe more accurate than theone estimated from Period04. Searching for only one frequency obtained by fitting the frequency and its harmonics. The of the highest amplitude at a time and prewhitening for it source of data is given in the last column of Table5. The andthenrepeatingtheproceduregavepractically thesame O−C diagram isshown in Fig.13. Thisplot indicatesthat results as searching for all theharmonics simultaneously. thepulsation period hasbeen constant duringthelast 40y.

See more

The list of books you might like

Upgrade Premium
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.
DMCA & Copyright: Dear all, most of the website is community built, users are uploading hundred of books everyday, which makes really hard for us to identify copyrighted material, please contact us if you want any material removed.
Copyright @ PDFdrive.to, 2022 | We love our users