TheAstronomicalJournal,131:1674–1686,2006March A #2006.TheAmericanAstronomicalSociety.Allrightsreserved.PrintedinU.S.A. A CATALOG OF SPECTROSCOPICALLY SELECTED CLOSE BINARY SYSTEMS FROM THE SLOAN DIGITAL SKY SURVEY DATA RELEASE FOUR Nicole M. Silvestri,1 Suzanne L. Hawley,1 Andrew A. West,1 Paula Szkody,1 John J. Bochanski,1 DanielJ.Eisenstein,2Peregrine McGehee,3,4 Gary D. Schmidt,2 J. Allyn Smith,5 Michael A. Wolfe,1 Hugh C. Harris,6 Scot J. Kleinman,7 James Liebert,2 Atsuko Nitta,7 J. C. Barentine,7 Howard J. Brewington,7 John Brinkmann,7 Michael Harvanek,7 Jurek Krzesin´ski,7,8 Dan Long,7 Eric H. Neilsen, Jr.,7,9 Donald P. Schneider,10 and Stephanie A. Snedden7 Received 2005 October 5; accepted 2005 October 27 ABSTRACT We present a spectroscopic sample of 747 detached close binary systems from the Sloan Digital Sky Survey (SDSS) Fourth Data Release. The majority of these binaries consist of a white dwarf primary and a low-mass secondary(typicallyMdwarf)companion. Wehavedeterminedthetemperatureandgravityfor 496ofthewhite dwarfprimariesandthespectraltypeandmagneticactivitypropertiesfor661ofthelow-masssecondaries.Wehave estimatedthedistancesforeachofthewhitedwarf–main-sequencestarbinariesandusewhitedwarfevolutionary grids to establish the age of each binary system from the white dwarf cooling times. With respect to a spectro- scopicallyidentifiedsampleof(cid:1)8000isolatedMdwarfstarsintheSDSS,theMdwarfsecondariesshowenhanced activitywithahigheractivefractionatagivenspectraltype.Thewhitedwarftemperaturesandgravitiesaresimilarto thedistributionof(cid:1)1900DAwhitedwarfsfromtheSDSS.Theagesofthebinariesinthisstudyrangefrom(cid:1)0.5Myr tonearly3Gyr(averageage(cid:1)0.20Gyr). Key words: binaries:close—novae,cataclysmicvariables—stars:activity—stars:low-mass,browndwarfs— whitedwarfs Online material: color figure, machine-readable tables 1. INTRODUCTION binarysystemsrepresentthelikelyprogenitorpopulationtomost cataclysmicvariables(CVs)andperhapsTypeIasupernovaeand Oneofthemanyinterestingby-productsoftheSloanDigital arethereforeasubjectofkeeninterest.Inaddition,theyprovidean SkySurvey(SDSS;Yorketal.2000)isthedetectionofalarge interestingtestofstellarevolutionforboththeprimaryandsec- number of unresolved, close binary star systems. In particular, ondarystarsinthebinaryenvironment.RecentSDSSsurveysof thefive-colorphotometryandspectracoveringnearlytheentire white dwarfs (Kleinman et al. 2004, hereafter K04; Eisenstein opticalwavelengthrangepermiteasyidentificationofunresolved etal.2005)andMdwarfs(Westetal.2004,hereafterW04)have binariesconsistingofa(blue)whitedwarfprimaryanda(red)sec- summarizedthebulkpropertiesofthesenearbystellarpopulations ondarystar,typicallyalow-massmain-sequenceMdwarf.These astheyevolveinisolation.Herewepresentalargesampleofwhite dwarf–Mdwarf(WDM)binariesidentifiedspectroscopicallyin theSDSSthroughtheFourthDataRelease(Adelman-McCarthy 1 DepartmentofAstronomy,UniversityofWashington,Box351580,Seattle, etal.2006,hereafterDR4).Wediscussthesampleselection,pro- WA98195;[email protected],[email protected],west@astro videtablesofthepropertiesoftheindividualcomponentsofeach .washington.edu,[email protected],[email protected], [email protected]. system(e.g., temperature,gravity, andage for the whitedwarfs; 2 DepartmentofAstronomyandStewardObservatory,UniversityofArizona, spectraltype,H(cid:1) emission,equivalent width,and spectroscopic Tucson, AZ 85721; [email protected], [email protected], parametersfortheMdwarfs),andcomparethepropertiestothose [email protected]. oftheisolatedwhitedwarf(K04)andMdwarfsurveys(W04).In 3 Los Alamos National Laboratory, LANSCE-8, MS H82, Los Alamos, NM87545;[email protected]. subsequent papers we will (1)explore the presenceof magnetic 4 DepartmentofAstronomy,NewMexicoStateUniversity,P.O.Box30001, whitedwarfsinthesesystemsandtheimplicationsfortheproduc- Department4500,LasCruces,NM88003;[email protected]. tionofmagneticCVsand(2)considerasubsetofthesamplefor 5 LosAlamosNationalLaboratory,ISR-4,MSD448,LosAlamos,NM87545; whichwehaveextensiveradialvelocity information (andhence [email protected]. 6 USNavalObservatory,P.O.Box1149,Flagstaff,AZ86002;[email protected] periods)tounderstandindetailthechangesthatoccurasthestars .mil. lose angular momentum and decrease their orbital separation, 7 New Mexico State University, Apache Point Observatory, P.O. Box 59, evolvingintoyetcloserbinarysystems,someontheirwaytothe Sunspot,NM88349;[email protected],[email protected],[email protected] CVphaseofevolution. .edu,[email protected],[email protected],[email protected], Mostpreviousattemptstoidentifyandstudywhitedwarfbi- [email protected],[email protected],[email protected],[email protected] .edu. narieswithlow-masscompanionshavebeenconfinedtowide, 8 Mount Suhoro Observatory, Cracow Pedagogical University, Ul. Pod- commonpropermotionsystems(seeLuyten1963;Oswaltetal. chorazych2,30-084Cracow,Poland;[email protected]. 1996;Silvestrietal.2001,2002)wheretheevolutionofthesys- 9 FermiNationalAcceleratorLaboratory,P.O.Box500,Batavia,IL60510; temproceedsasforsinglestarsduetothelarge((cid:1)103AU)orbital [email protected]. 10 DepartmentofAstronomyandAstrophysics,PennsylvaniaStateUniver- separationofthecomponents.Studiesofpre-CVssuchasSchultz sity,UniversityPark,PA16802;[email protected]. et al. (1996), Vennes et al. (1999), Marsh (2000), Hillwig et al. 1674 Report Documentation Page Form Approved OMB No. 0704-0188 Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number. 1. REPORT DATE 2. REPORT TYPE 3. DATES COVERED 2006 N/A - 4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER A Catalog of Spectroscopically Selected Close Binary Systems From the 5b. GRANT NUMBER Sloan Digital Sky Survey Data Release Four 5c. PROGRAM ELEMENT NUMBER 6. AUTHOR(S) 5d. PROJECT NUMBER 5e. TASK NUMBER 5f. WORK UNIT NUMBER 7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION Library U.S. Naval Observatory 3450 Massachusetts Avenue, N.W. REPORT NUMBER Washington, DC 20392-5420 9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S) 11. SPONSOR/MONITOR’S REPORT NUMBER(S) 12. DISTRIBUTION/AVAILABILITY STATEMENT Approved for public release, distribution unlimited 13. SUPPLEMENTARY NOTES 14. ABSTRACT 15. SUBJECT TERMS 16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF 18. NUMBER 19a. NAME OF ABSTRACT OF PAGES RESPONSIBLE PERSON a. REPORT b. ABSTRACT c. THIS PAGE SAR 13 unclassified unclassified unclassified Standard Form 298 (Rev. 8-98) Prescribed by ANSI Std Z39-18 CLOSE BINARY SYSTEMS 1675 (2000), Maxted etal. (2004), and mostrecentlyMorales-Rueda The color cuts were later refined [ðu(cid:2)gÞ(cid:2)1:314ðg(cid:2)rÞ< etal.(2005)andDobbieetal.(2005)involvetoofewsystemsfor 0:61; ðr(cid:2)iÞ>0:91; ði(cid:2)zÞ>0:49] to improve the selection robuststatisticalstudy.Raymondetal.(2003)carriedoutapre- ofthesepairs(P.McGehee2005,privatecommunication).While liminarysurveyof WDMsystemsintheSDSSusingdatafrom thistargetingalgorithmisrelativelyefficient,itisoflowpriorityin theEarlyDataRelease(Stoughtonetal.2002,hereafterEDR)and theoverallSDSSspectroscopictargetingpipeline,andtheresult- Data Release One (Abazajian et al. 2003). Their sample com- ing spectra obtained during normal survey operations represent prised109systems,ofwhich106arecontainedinoursample(the onlyasmallfraction(<20%)ofourSDSSspectroscopicsample. remainingthreehavebeenreclassifiedanddonotfitoursample Thecolorsofwhitedwarf–dominatedunresolvedpairsaresimilar criteria).Ourmuchlargersampleandimprovedanalysisenables to many other interesting cosmological objects targeted by the ustoextendand refine their results.We anticipatethatapproxi- SDSS;hence,themajorityofsystemsinthissamplewereserendip- mately500additionalsystemswillbeaddedtothesamplebythe itouslytargetedforspectroscopybyotherhigherprioritypipelines endoftheSDSS.ThenewSloanExtensionforGalacticUnder- (e.g.,quasars,whitedwarfs).Therefore,ourspectroscopiccompi- standingandExploration(SEGUE)iscontinuingtotargetthese lationofclosebinarysystemsisnotwelldefinedphotometrically, objects and will likely discover more of these close binaries at norisitstatisticallycomplete. lowerGalacticlatitudes. Thesimilarityincolortosomanydifferentobjectstargetedby Thepaperproceedsasfollows.Inx2wedescribethesample theSDSSmakesthetaskofidentifyingtheseobjectsbasedsolely selectionandpresentextensivetablesgivingthefullphotometric on their ugriz photometryquite difficult.Typicalcolor cuts (see and spectroscopic description of the systems. In x 3 we derive Smolcˇic´ etal.2004)donotsingleoutthepairsinwhichtheflux basicparametersforthewhitedwarfprimariesandcomparethese isveryblue(dominatedbythefluxfromthewhitedwarf)orvery with the extensive survey of white dwarfs in the SDSS (K04). red(dominatedbythefluxfromtheMdwarf).Thefinalcolorcuts Section4containsadiscussionoftheMdwarfsecondariesand for the spectroscopic targeting algorithm were necessarily quite comparisontotheSDSSsurveyofMdwarfsinW04.Section5 narrowinordertoavoidseverecontaminationbyisolatedwhite summarizesourconclusionsandplansforfuturework. dwarfs and by M dwarfs from the low-mass end of the stellar locus.SearchingtheSDSSspectroscopicdatabaseonlyforobjects 2. WDM SAMPLE SELECTION withthisstringentcolorselectionwouldexcludemanysystems. Thisstudyrepresentsacompilationof(cid:1)750closebinarysys- Instead,weresortedtoaninitialvisualinspectionofeachof temsobservedbytheSDSSthroughtheDR4.TheSDSS(York the640fibersfromeachofthe(cid:1)1000DR4platesinaneffortto etal.2000;EDR;Abazajianetal.2003,2004,2005;Adelman- recover more spectroscopically observed pairs. The obviously McCarthyetal.2006)isanimagingandspectroscopicsurveyof time-consumingnatureofthisprocesspromptedthecreationof 104 deg2 of sky in five colors (u, g, r, i, and z; Fukugita et al. analgorithmthatweusedtoautomaticallysearcheachspectrum 1996;Gunnetal.1998,2005).Detailsontheimageprocessing foravarietyofhydrogen(BalmerH(cid:1),H(cid:2),H(cid:3))andheliumab- andcalibrationfortheSDSScanbefoundinFan(1999),Lupton sorptionfeaturesthatidentifywhitedwarfs.Wealsosearchedfor etal.(1999),Hoggetal.(2001),Smithetal.(2002),Pieretal. emission features to find active M dwarfs and CVs. Each se- (2003), Ivezic´ et al. (2004), and Tucker et al. (2005). Objects lectedspectrumischeckedbyeye,andifitisabonafideWDM selectedforfollow-upspectroscopyareobtainedwithtwinfiber- pair,eachbinarycomponentisassignedacoarseidentification. fed double spectrographs, which cover a wavelength range of The WDM pairs with low signal-to-noise ratio (S/N<5) or 3800to(cid:1)10,0008. non-DA/DB primaries are missed by our algorithm, but many We restrict the sample to systems that have been observed havebeenrecoveredthroughthewhitedwarfidentificationpipe- spectroscopically, are detached (no observational signature of linedescribedinK04andEisensteinetal.(2005).Therefore,we mass transfer or an accretion disk), contain a white dwarf pri- believeoursamplecomprisesalmostallofthespectroscopically mary,11andareinanunresolvedbinary12withalow-mass,non- observedWDM binariesinthe SDSSthrough the DR4 datare- degeneratecompanion.AnexamplespectrumofaWDMsystem lease(2005July). isshowninFigure1.ThetoppaneldisplaysanSDSSspectrum Table1liststheSDSSphotometry(psfmag)fortheentiresam- ofanunresolved,closebinarysystemconsistingofablue,white ple,including106ofthe109pairsfromRaymondetal.(2003). dwarfprimaryandared,Mdwarfsecondary.Theobservedspec- Columns(1)–(4)givetheSDSScoordinatename(hhmmss.ss(cid:3) trumisthesuperpositionoftheunresolvedspectralenergydis- ddmmss.s J2000.0), plate number, fiber number, and Modified tributionsofthetwocomponents.Thebottompanels(described JulianDate(MJD)oftheobservation,respectively.Column(5) inmoredetailinx2.1)showthespectraoftherecoveredwhite lists the coarse identifications of each component of the pair; dwarf(left)andtherecoveredMdwarf(right).Notethechromo- columns(6)and(7)listtherightascension(J2000.0indecimal sphericemissionfromthemagneticallyactiveMdwarfseenin hr) and declination of the object (J2000.0 in decimal degrees); theH(cid:1)((cid:1)65638)lineinthebottomrightpanel.TheMdwarf columns(8)–(22)givetheugrizmagnitudes,errors,andGalactic H(cid:1)emissionmayalsobeseensuperposedontheH(cid:1)absorption extinctions;column(23)liststheoriginaldatarelease;andcol- lineofthewhitedwarfinthetoppanel. umn(24)listsadditionalnotes.Wefollowtheconventionofde- AsdescribedinRaymondetal.(2003),aninitialsetofpho- signatinguncertainspectraltypeclassificationwithacolon;i.e., tometric selection criteria was developed to identify these sys- any spectral type followed immediately by a colon indicates temsusingSDSSphotometry.Fourcolors(u(cid:2)g<0:45[blue], uncertaintyintheby-eyeclassificationoftheobject.Forexam- g(cid:2)r<0:70[blue],r(cid:2)i>0:30[red],andi(cid:2)z>0:40[red]) ple,DA:+dMindicatesthatthemostlikelyspectraltypeforthe yield reliable identification to a limiting magnitude of g<20. whitedwarfprimaryisDA,althoughforvariousreasons(most commonlylowS/N)theclassificationshouldbetakenastentative untilfollow-upspectroscopycanbeperformed.Wemaintainthis 11 Wedefinethewhitedwarf,theoriginalhigh-masscomponent,asthepri- notationinalltables.Itshouldbenotedthattheclassificationsin maryinthesesystems,unlikewidebinarysystemswheretheprimaryistypically thispapersupersedethemoregeneralclassificationspresentedin definedasthebrighterofthetwocomponents. 12 Aparallelinvestigationofcommonpropermotionbinarysystemsinthe Eisensteinetal.(2005).Updatedclassificationsfortheseobjects SDSSisunderway,andwereferthereadertoSmithetal.(2004)fordetails. arepresentedinxx3and4(Tables4and5). 1676 SILVESTRI ETAL. Fig.1.—Top: Spectrum of the unresolved close binary systemSDSSJ113655.17+040952.6; R(cid:1)1800. Bottom left: Spectrum of therecovered white dwarf. Bottomright:RecoveredMdwarf. Updated photometry and spectroscopy from SDSS data re- photometrically by Smolcˇic´ et al. (2004) are recovered spectro- leasesbeyondtheEDRledtothereclassificationofthreeofthe scopically,mostlikelybecausetheyarenottargetedwiththesame 109objectsfromtheRaymondetal.(2003)sample.Newpho- frequencyasthebluerWDMsthatmorecloselymatchthecolors tometry and tentative spectral types are given in Table 1 for ofhigh-prioritytargetsinthemainSDSS.Smolcˇic´ etal.(2004) SDSSJ143947.62(cid:2)010606.9(DA+:e)andSDSSJ234459.62+ used strictcolor cuts toavoid contamination of their sample by 002749.9(DA+:).TheobjectSDSSJ113722.25+014848.6isthe whitedwarfsandquasars.ThecutsaremostobviousinFigure3 CVRZLeo;detailscanbefoundinSzkodyetal.(2003).Asthis andaccountfortheabruptterminationofthesample(dots)above systemisundergoingmasstransfer,weexcludeitfromoursample. theDAwhitedwarfsinthefigure.Thespectroscopicallyidentified Figures2and3showtheugrandgricolor-colordiagramsfor sampleof WDMsfillsthegapbetweenthephotometricallyiden- thespectroscopicallyidentifiedWDMs(opencirclesinbothfig- tifiedsamplesofwhitedwarfsandWDMs,recoveringmanyob- ures), as compared to the photometrically identified WDMs of jects that would have been missed by the photometric cuts for Smolcˇic´ etal.(2004;dotsinbothfigures)andthespectroscop- eithersurvey. ically identified white dwarfs of K04 (crosses in both figures). 2.1. Separating the Components of the WDM Systems Thecolorsarethepoint-spreadfunction(PSF)magnitudesfrom the latest version of the photometry pipeline (photo 5.4.25; Toinvestigatethepropertiesoftheindividualcomponentsof Luptonetal.2001),asdescribedbyFukugitaetal.(1996),Gunn theWDMbinarysystems,weneedarobustmethodofseparat- etal.(1998),Hoggetal.(2001),Smithetal.(2002),andEDR. ingthetwostellarspectraforindividualanalysis.Ourmethoduses TheyhavebeencorrectedforGalacticextinction(Schlegeletal. MstartemplatesandcolorindicesasdescribedinHawleyetal. 1998);thiscorrectionisusuallyminorfortheserelativelynearby (2002) and the white dwarf model atmospheres of D. Koester, systems.ThecontributionfromtheMdwarfcompanionskews describedinK04andEisensteinetal.(2005).Weuseaniterative thelocationofthepairsincolor-colorspacetoreddercolorsin (cid:4)2 methodtofit and subtractwhite dwarf models and M dwarf bothfiguresfromthelocusoccupiedbyisolatedDAwhitedwarfs. templatesfromeachWDMspectrumtofindthebestfitmodeland Veryfewofthered(Mdwarf–dominated)WDMsthatareseen templateforthewhitedwarfandMdwarf,respectively. TABLE1 SDSSWhiteDwarf+Main-SequenceStarBinaries Name SpectralTypea R.A.b Decl. (SDSSJ) Plate Fiber MJD (SP1+SP2) (deg) (deg) u (cid:5) A g (cid:5) A r (cid:5) A i (cid:5) A z (cid:5) A DataReleasec Notesd psf u u psf g g psf r r psf i i psf z z (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) 001029.87+003126.2...... 0388 545 51793 DZ:+dM 2.62448 00.52396 21.93 0.19 0.14 20.85 0.04 0.10 19.98 0.03 0.08 19.00 0.02 0.06 18.42 0.04 0.04 EDR 001324.33(cid:2)085021.4..... 0652 321 52138 WD:+dM 3.35139 (cid:2)08.83929 19.69 0.05 0.16 19.73 0.03 0.12 19.67 0.03 0.09 19.08 0.03 0.07 18.57 0.04 0.05 DR1 001726.63(cid:2)002451.2..... 0687 153 52518 DA+dMe 4.36099 (cid:2)00.41422 19.68 0.04 0.14 19.29 0.03 0.10 19.03 0.02 0.07 18.19 0.02 0.06 17.54 0.03 0.04 R 001733.59+004030.4...... 0389 614 51795 DA+dM 4.38996 00.67511 22.10 0.40 0.13 20.79 0.14 0.10 19.59 0.03 0.07 18.17 0.02 0.05 17.39 0.02 0.04 EDR/R 001749.24(cid:2)000955.3..... 0389 112 51795 DA+dMe 4.45519 (cid:2)00.16539 16.57 0.02 0.13 16.87 0.02 0.10 17.03 0.01 0.07 16.78 0.01 0.05 16.47 0.02 0.04 EDR/R Notes.—Table1ispublishedinitsentiretyintheelectroniceditionoftheAstronomicalJournal.Aportionisshownhereforguidanceregardingitsformandcontent. a Spectraltype;initialidentification;e:emissiondetectedbyeye. b Rightascension anddeclinationareJ2000.0equinox. c (R)OriginallypublishedinRaymondetal.(2003);(K)originallypublishedinK04.Numbersinparenthesesareplatenumbersfortheobject’soriginaldatarelease. d Doubleasterisksindicatepotentiallow-gravity(logg<7)whitedwarfs. 1678 SILVESTRI ETAL. Vol. 131 TABLE2 SDSSWhiteDwarfSpectralTypes SpectralType NumberofComponents DA(B,H)................................. 666(6,3) DB(A)........................................ 26(2) DC............................................. 16 DZ(A)........................................ 5(1) DO............................................. 1 DQ............................................. 1 WDa........................................... 31 Notes.—Atmosphericcompositionclassification.Refertox3.1 andTable5forspectralsubtypes.Lettersandnumbersinparen- thesesindicatemultipleatmosphericcomponents.Forexample, DB(A)and26(2)indicatethat26whitedwarfsareclassifiedas DBandtwoareofmixedatmospheretypeDBA. a WD:unclassified.Detectedthepresenceofahot,bluecom- panionbutcouldnotclassifythespectraltype.Refertox3for details. analyzedindependently,asdescribedinx3(forthewhitedwarfs) andx4(fortheMdwarfs). Fig.2.—Theugrcolor-colordiagramfortheWDMsinoursample.Thespec- 3. THE WHITE DWARF PRIMARIES troscopicallyidentifiedWDMsareplottedasopencircles(left).Forcomparison, Theoverwhelmingmajorityofthewhitedwarfprimariesare intherightpanelweplotthephotometricallyidentifiedWDMsofSmolcˇic´etal. (2004;dots)andtheisolatedwhitedwarfsofK04(crosses). hydrogen-atmosphere,DAwhitedwarfs.Thehandfulofnon-DA primariesareoftypesDB,DC,DZ,DO,andDBA.Table2lists the number of white dwarfs of each spectral type in the cur- AnexampleofthissubtractionprocessisintroducedinFig- rentsample(seeTable3forthebreakdownofMdwarfspectral ure1.TheWDM,SDSSJ113655.17+040952.6(top),wassep- types). aratedintoitstwocomponentstarsbysubtractingabest-fitmodel FollowingtheprescriptionoutlinedinK04andimplementing DAwhite dwarf with TeA(cid:1)13;000 K and logg¼8:0 from the theirautofitprogram,thespectraarefittedtothetemperature– original spectrum to reveal the M dwarf. A best-fit M5 template surface gravity grid of theoretical pure hydrogen and pure he- wasthensubtractedfromtheoriginalspectrumtorevealthewhite liumwhitedwarfmodelspectraofD.Koester(seeFinleyetal. dwarf.Theprocessworksquitewellwithminimalsmoothing [1997]andK04forafulldescriptionofthemodelsandautofit (maximumboxcarof78forspectrawithS/N>10).Thesep- program), with 6000 K<TeA<100;000 K and logg from 5 aratedwhitedwarfandMdwarfspectraarethenprocessedand to 9. Our white dwarf spectral type identifications follow the convention of Sion et al. (1983). An example of one of our TABLE3 SDSSMDwarfSpectralTypes SpectralType NumberofComponents K0........................................... 1 K5........................................... 4 K7........................................... 27 M0.......................................... 64 M1.......................................... 42 M2.......................................... 81 M3.......................................... 149 M4.......................................... 179 M5.......................................... 66 M6.......................................... 33 M7.......................................... 8 M8.......................................... 2 M9/L0.................................... 5 M:a.......................................... 67 C............................................. 3 :b............................................. 15 Notes.—Finalclassification.Refertox4fordetails. a M:Unclassified.Detectedthepresenceofaredcompanion Fig.3.—Thegricolor-colordiagramfortheWDMsinoursample.Thespec- butcouldnotclassifythespectraltype.Refertox4fordetails. troscopicallyidentifiedWDMsareplottedasopencircles(left).Forcomparison, b Redexcessdetectedinphotometryand/oremissiondetected intherightpanelweplotthephotometricallyidentifiedWDMsofSmolcˇic´etal. inWDspectra.Suspectpresenceofacompanionbutneedspec- (2004;dots)andtheisolatedwhitedwarfsofK04(crosses). troscopicfollow-uptoconfirm. No. 3, 2006 CLOSE BINARY SYSTEMS 1679 tendstobemostcontaminatedbytheMdwarfandisthefea- tureleastlikelytoberecoveredinthesubtractionprocess.Tests onisolatedwhitedwarfsfromtheK04studythatexcludeH(cid:1) from the fit show very little change in the measured T and eff loggdeterminedbytheautofitprogram.ThechangeinT eff and logg was within the measurement errors for all but the lowestS/Nobjects(<10).Wehaveexcludedallobjectsbelow this S/N threshold from our final analysis to ensure we are using only the most reliable values of T and logg. We have eff foundthatthepoorestfitstothewhitedwarfspectraresultfrom lowS/N(<10),notresidualMstarcontributionsfromthesub- tractionprocess. Wewereabletoaccuratelyfitmodelsto496ofthe747bina- riesinoursample.Modelfitscouldnotbeobtainedfor there- mainingwhitedwarfsduetothelowS/Nofthespectrumand/or a largeM dwarfflux contribution that could not be adequately removedfromthespectrum.Wemadenoattempttodetermine the T and logg of white dwarfs with spectral types DC, DZ, eff DO,orDBA.Table4liststhewhitedwarfparametersobtained fromtheatmosphericmodelfits(T ,logg,andtheDAandDB eff subtype),aswellasthecoolingagesfromevolutionarymodels ofBergeronetal.(1995). Fig.4.—WDMSDSSJ172406.14+562003.1fromoursample.D.Koester’s Figure 5 plots the effective temperature distribution for the DAwhitedwarfatmosphericmodelfittothespectrumisshownwithreddening separated DAwhite dwarf primaries, and Figure 6 shows the (solidline)andwithoutrefluxing/reddening(dashedline).Thewhitedwarfhas loggdistribution.TheaverageT ofthe483DAwhitedwarfs eff TeA¼36;250Kandlogg¼7:2,withafinalspectraltypeofDA1.4.Notethe (weexcludethe13DBdwarfsfromthesefiguresforcomparison excellentfitevenwiththestrongH(cid:1)emissionandresidualfeaturesinthered with K04) is (cid:1)21,200 K. The temperature distribution of the fromthecompanionMdwarf,whichwerenotperfectlysubtractedfromthe spectrum.[SeetheelectroniceditionoftheJournalforacolorversionofthis WDMs in Figure 5 is broad, reflecting the large range of Teff figure.] foroursample.TheK04distributionisstronglypeakedatlow T (<10,000K),withahigh-temperaturetail.Ourdistribution eff fits from the autofit program is given in Figure 4. The top hasasimilarhigh-temperaturetailbutmanymorewhitedwarfs paneldisplaysthelikelihoodcontours(1,2,3,5,and10(cid:5))in around 20,000 K. This is most likely a selection effect in the logg-subtype space, where the subtype represents the T of spectroscopictargetingoftheWDMobjects.Asdescribedabove, eff thewhitedwarfinunitsof50;400/TeA (Sionetal.1983).The manyofourobjectsweretargetedbythequasarpipeline,which middleandbottompanelsplotthefullSDSSspectrumandthe selects many blue objects. The flux from an object selected by regionsofthewhitedwarfspectrumwhereinterestingfeatures thequasarpipelineisdominatedbythewhitedwarf.Foranaver- aretypicallylocated(e.g.,Balmerhydrogenandheliumabsorp- ageMstarcompanion,thewhitedwarfthushastobequitehot. tion). The solid line is the fit that includes reddening, and the Systemswithcoolerwhitedwarfsarenottargetedasfrequently dashed line is the fit without reddening/refluxing (see K04 for (seex2discussion)and/orarenotrecognizedasspectroscopic details). pairs because the M dwarf masks the flux from the cool white TheSDSSPSFmagnitudesoftheWDMpairsareusedinthe dwarfcompanion. overallfit,asforthesinglewhitedwarfsdescribedinK04.We We are conducting a parallel study in which we have al- makenoefforttodeconvolvethewhitedwarf’scolorcontribu- readyperformedradialvelocityfollow-uponover30ofthese tion from the M dwarf’s color contribution. The fits are quite systems to determine their orbital periods. Preliminary results goodifthecontinuumcontributionfromtheMdwarfandcon- suggest that several of these systems have orbital periods that taminationoftheH(cid:1)absorptionfeaturearesmall(oreffectively placethemwithintheCVperiodgap(2hr<P <3hr).Aseach orb removed) and the spectrum is of high quality (S/N(cid:4)15). In systemmovesintoRochelobecontactandbeginsmasstransfer, some cases, residual M dwarfflux (as seen in Fig. 4 between thewhitedwarfisheatedbytheaccretionprocess,therebyde- (cid:1)7000 and 9000 8) remains in the spectrum. Our fits to the stroyinganyageinformationfromthecoolingofthepreviously spectrawereappliedbetween3500and54008.Thisexcludes detachedwhitedwarf.Withoutknowingtheorbitalperiodofthe boththe(cid:1)7000–90008regiondiscussedaboveandH(cid:1),which hot systems, we must be cognizant of the fact that some small TABLE4 WhiteDwarfParameters Name SpectralType WD MSSpectral T (cid:5) Age eff TeA (SDSSJ) Plate Fiber MJD (SP1+SP2) Subtype (cid:5) Type log(g) (cid:5) (K) (K) (yr) Notes WDSubtype logg 001029.87+003126.2......... 0388 545 51793 DZ:+dM ... ... M0 ... ... ... ... ... 001324.33(cid:2)085021.4........ 0652 321 52138 WD:+dM ... ... M4 ... ... ... ... ... 001726.63(cid:2)002451.2........ 0687 153 52518 DA+dMe 3.1 0.1 M4 8.1 0.1 16507 532 1.84E+08 001733.59+004030.4......... 0389 614 51795 DA+dM 2.5 0.2 M4 7.7 0.2 19888 1591 4.82E+07 001749.24(cid:2)000955.3........ 0389 112 51795 DA+dMe 0.7 0.0 M2 7.9 0.1 75044 0 7.10E+05 Notes.—Table4ispublishedinitsentiretyintheelectroniceditionoftheAstronomicalJournal.Aportionisshownhereforguidanceregardingitsformandcontent. 1680 SILVESTRI ETAL. Vol. 131 Fig.5.—TeffdistributionoftheDAwhitedwarfsinSDSSclosebinarysys- Fig. 7.—White dwarf Tefffrom D. Koester’s white dwarf models vs. the tems(top)ascomparedtotheisolatedSDSSDAwhitedwarfsofK04(bottom). whitedwarfagefromthecoolingmodelsofP.Bergeron(seeBergeronetal. TheTeffdistributionofwhitedwarfsinclosebinariesisbroaderthanintheisolated 1995).Theinsetpanelgivesthecoolingagedistributionoftheclosebinaries.The DAwhitedwarfsample.Therearemanyhotwhitedwarfsinthesesystems,indi- majorityofthemeasuredwhitedwarfsarerelativelyyoung(age<2;108yr). catingthatalargefractionofthesampleisyoungerthantheisolatedDAs.Seex3 foranalternativeexplanation. numberofthesesystemsmayhaveundergoneaperiodofmass transferthatreheatedthewhitedwarf.Inthesecases,theages determinedfromthemodelfitscanbeconsiderablyyoungerthan theactualageofthesystem. Employingtheevolutionarygrids(updatedforSDSScolors) of Bergeron et al. (1995), we determine the age via bicubic interpolation through the evolutionary models for the T and eff logg of each white dwarf. Figure 7 plots the T versus age eff forthewhitedwarfsinoursample.Linesofconstantloggare given forlogg¼7:0, 7.5, 8.0, 8.5, and 9.0. The histogramin the upper right corner displays the resulting white dwarf age distribution. WefindthattheaveragecoolingageofDAwhitedwarfswith TeA(cid:1)21;200 K is approximately 0.20 Gyr. Compared to the average age impliedby thenumerouscool white dwarfsinthe K04sampleandtheageoftheGalacticdisk((cid:1)7–9Gyr;Liebert etal.1988;Oswaltetal.1996),thesesystemsarequiteyoung. Theyare,however,olderthanthe30well-observedclosebinary systemsdescribedinSchreiber&Ga¨nsicke(2003),whichsug- gests that the SDSS is finding cooler white dwarfs in close bi- nariesthanearliersearchesforthesesystemshaverevealed. The autofit program returned 20 white dwarfs with low- gravity (logg<7) fits, which may indicate that they are quite lowmass(M (cid:5)0:45M ).Mostoftheseresultedfrompoor WD (cid:6) fitsasaresultofinadequateMdwarffluxsubtractioncombined withlowS/N.However,fiveoftheselow-gravityobjectsappear tohavereasonablefitsbutareveryfaint(g(cid:4)20),andasaresult have very low S/N.13 Low-mass white dwarfs in close binary systems are integral to our understanding of the mass loss and diffusion episodes in the stages leading up to the formation of Fig.6.—PlotoftheloggdistributionoftheDAwhitedwarfsinSDSSclose whitedwarfs(Benvenuto&DeVito2004;Althausetal.2004). binarysystems(top)ascomparedtotheisolatedSDSSwhitedwarfsofK04 Ithasbeensuggestedthatalow-masswhitedwarfcanonlybe (bottom).Thereappearstobeabroaderrangeoflogg(masses)associatedwith close pairs than the relatively narrow distribution of loggfor the field white dwarfsample,butfewerathighlogg.Thehighloggtailismostlikelyanarti- 13 TheSDSShasuncoveredavarietyofextremelowmasswhitedwarfs.See factofthemodelfitsatTeA<10;000K.SeeK04fordetailsontheirhighlogg K04,Liebertetal.(2004),andEisensteinetal.(2005)fordetailsontheseun- tail. usuallylowsurfacegravityobjects. No. 3, 2006 CLOSE BINARY SYSTEMS 1681 The spectra of these systems may seem to resemble low ac- cretionrateCVs(seeSzkodyetal.2004,2005),butwecansay that the emission seen in these systems is not a result of mass transfer. Emission line widths in CVs, even low M˙ CVs, are (cid:1)30–40 8 or more at their base near the continuum in H(cid:1). In contrast, all of our systems have line widths <20 8 at the continuum. Further, many of the emission lines in CVs have structurethatindicatesthepresenceofanaccretiondisk,andthe hydrogen emission lines are often accompanied by He I and HeIIlinesinemission.Wedetectnoneofthesesignaturesinthe 27 objects presented here, with the exception of the two sys- tems discussed below. Instead, the hydrogen emission features weobservearetypicalofemissionlinesproducedbymagneti- callyactiveMdwarfstarswithstrongchromosphericemission, or by M dwarfs that are subject to significant irradiation that causesexcessphotoionizationandsubsequentrecombinationin thehydrogenBalmerlines. Theemissionweobservemaybesomecombinationofboth of these processes for most of the systems in our subsample. However, two of the 27 objects (SDSS J131751.72+673159.4 andSDSSJ143947.62(cid:2)010606.9)showHeIinemissioninad- ditiontotheemissioninalloftheobservableBalmerserieslines (seeFig.8,top).ThesimultaneouspresenceofHandHeIlines Fig. 8.—Three white dwarfs with emission in H(cid:1), H(cid:2), and H(cid:3) and no isoften producedbyirradiation ofanMdwarfsecondarybya obviousopticalcounterpartinthered.Thereare24additionalwhitedwarfswith similarstrongemissionandnoopticalcounterpart.Thefeaturenear56008in hotprimarystar,whileHeIisnotusuallyobservedintheemis- eachofthespectraisanartifactoftheSDSSdatareductionsandisnotareal sionsofnormalmagneticallyactiveMdwarfs.Thewhitedwarfs feature. inthesesystemsarethehottestinthissubsample,makingirra- diationapromisingcandidateforproducingtheobservedemis- formed in the presence of a binary companion (Marsh et al. sioninthesetwoobjects. 1995);alargesampleofthesesystemswouldbeaninteresting Of particular interest is SDSS 121209.31+013627.7. The test bed for a variety of theories. If these systems do in fact spectrumofthisobject,firstdiscoveredbySchmidtetal.(2003), harbor low-mass white dwarfs, they would nearly double the isofamagneticwhitedwarfwitha13MGfield.Theemission known population of such systems. We list the fit parameters feature at H(cid:1) is weak and not obvious in the SDSS discovery forthesesevenobjectsinTable1buthavenotincludedthemin spectrum.ThestrongH(cid:1)emissionwasconfirmedwithfollow- anyofthefigurespending additionalspectroscopytoconfirm upspectroscopyatApachePointObservatory14in2004March. theirnature.Thesesevenwhitedwarfsareidentifiedincolumn Thisappearstobethefirstobservationofamagneticwhitedwarf (24)ofTable1. withanondegeneratebinarycompanion(seeLiebertetal.2005 The logg distributions of this sample and the K04 DAs are fordetails).Weincludethisobjectinourtablesforcompleteness quitesimilar,withafairlysymmetricdistributionaroundapeak andreferthereadertoSchmidtetal.(2005)foradetailedanal- of 7.9, as demonstrated in Figure 6. However, there are more ysisofthissystem. white dwarfs with high logg in the K04 distribution (bottom). Weattemptedtomatchall27whitedwarfswithemissionto It is probable these high-mass white dwarfs also exist in close objectsobservedintheTwoMicronAllSkySurvey(2MASS). binarysystems,butagain,thelackofsuchsystemsinoursample Only eight of the 27 objects were sufficiently bright to be ob- islikelyaresultofourinabilitytodistinguishthemfromisolated served in all three 2MASS colors (JHK ). Following the pre- S M stars. High-gravity white dwarfs have smaller radii. Paired scriptionoutlinedinWachteretal.(2003)andWellhouseetal. with an average M dwarf companion, the white dwarf would (2004),weplotthenear-IRcolorsoftheeightsystemsinFigure9 contributeverylittlefluxtothecombinedspectrumandwould and divide the plot into the same regions as Wellhouse et al. be very difficult to observe. Many such objects have probably (2004)basedonsimulatednear-IRcolors.RegionIistheloca- goneundetectedinSDSSspectroscopicstudiesandwillonlybe tionofsinglewhitedwarfs,whileregionIIcontainsWDMbina- revealedwithUVfollow-upobservations. riesandregionIIIiscomprisedofwhitedwarf–Ldwarfsystems. ObjectsinregionIVmaybeCVs,magneticwhitedwarfs,orwhite 3.1. White Dwarfs with Emission dwarfswithunusualatmosphericcompositions. Twenty-seven objects identified in this study (DA+:e) have AsdemonstratedinFigure9,sixoftheeightobjectsliewithin whatappeartobenormal,isolatedwhitedwarfspectralfeatures regions II and III with colors indicative of unresolved white with(insomecases,strong)hydrogenBalmeremissionfeatures dwarf–M dwarf or –L dwarf binary systems. The two objects in the spectra. Figure 8 shows three examples. Although there in region I clearly show emission at H(cid:1), H(cid:2), and H(cid:3), and the isnoobviousopticalsignatureofaclose,low-masscompanion uncertaintyintheircolorssuggeststhattheseobjectsmaybelong otherthanthepresenceoftheemissionatBalmerwavelengths, inregionIII. the source of the emission is probably a late M dwarf (or per- H(cid:1) emission has been observed in L dwarfs by groups such hapsLdwarf)companion.Mostofthe27whitedwarfsarehot asLiebertetal.(2003).Theyarguethatstrongflareeventswith (>20,000K)andbright.WhileMdwarfs aremuch largerthan a magnetic origin are common among these low-mass objects, white dwarfs, their relatively cool temperatures result in very lowluminosityatopticalwavelengths,sohotwhitedwarfswill 14 TheApachePointObservatory3.5mtelescopeisownedandoperatedby easilydominatethecombinedspectrum. theAstrophysicalResearchConsortium. 1682 SILVESTRI ETAL. Vol. 131 Fig.9.—Color-colorplotofeightofthe27DAWDswithBalmeremission, Fig. 10.—Distribution of M dwarf spectral types in WDM systems (top) recoveredin2MASSJHK.ObjectsinregionsIIandIIIhavecolorsindicative and in W04 (bottom). The negative spectral type designations indicate the ofanunresolvedlow-massScompanion. spectraltypesofthe32Kdwarfsecondariesinthesample.TheWDMMdwarf distributionisnarrowerandmoresymmetricthantheW04distribution.Seex4 fordetails. althoughthemechanismforproducingsustainedemissionismuch lessefficientthaninearliertypeMdwarfs(W04).Nevertheless, help to identify white dwarf binaries with very late type com- someLdwarfsdoshowpersistentH(cid:1)emission.Thus,thelow- panionsasdiscussedinx3.1. masscompanionsinregionIIImayindeedbeLdwarfs. Theselectioneffectsassociatedwiththelow-masssecondar- ies become apparent when compared with the M dwarf distri- 4. THE LOW-MASS SECONDARIES butionof W04.TheW04studyincludes7831spectroscopically The vast majority of the secondaries in this catalog are identified M dwarf stars. As displayed in Figure 10, the W04 Mdwarfstars(32areKdwarfs[dK],threearedwarfcarbonstars distribution is quite broad compared to the WDM distribution. [dC],and15cannotbeidentified[:]).Theassignmentofspectral TheW04sampleincludesmanymoreearlyandlateMdwarfs, typeisaccomplishedviatwotypingmethods:(1)aleast-squares whereastheseobjectswerelikelyselectedagainstintheWDM fit to template spectra of type G, K, M0–M9, and L0–L4 and sample.TheW04studyconcentratedonlyonMdwarfs,which (2)aweightedmeanofmolecularbandindicesasdescribedin accountsforthesharpcutoffatzero(M0)intheirdistribution. W04. Using these methods to predictthe best-fit spectral type, As with our sample, given the uncertainty in spectral type it each spectrum is visually inspected and the predicted spectral is possible that some of the M0 dwarfs in their study are late typeupdatedasnecessary.Theuncertaintyinthismethodisof Kdwarfs. order(cid:3)1spectralsubtypefortheMandLdwarfs.Wealsodis- tinguishbetweenearlyKdwarfs(K0),mid-Kdwarfs(K5),and 4.1.M Dwarf Activity lateKdwarfs(K7). Collisionally induced emission at hydrogen Balmer wave- The spectral type distribution of Figure 10 is symmetric, lengths (H(cid:1), H(cid:2), etc.) is formed in the magnetically heated narrow,andpeakedatM4((cid:2)1,(cid:2)3,and(cid:2)8valuesinthefigure chromosphere ofthe star. The H(cid:1)equivalentwidth is usedto representK7,K5,andK0dwarfs,respectively).Thisrepresents characterizemagneticactivityinMdwarfs(Hawleyetal.1996). aconvolutionoftheincreasingmassfunctiontowardlowermass WeusethetrapezoidalintegrationtechniqueoutlinedinW04to togetherwiththedecreasingvisibilityoflowermassstarscom- sumthefluxinH(cid:1)undertheemissionfeature.Welabelastaras paredtothewhitedwarfprimariesintheoptical.Thus,itprob- activeifithasanH(cid:1)equivalentwidthgreaterthan1.08.15The ablydoesnotgiveatruepictureofthesecondarymassfunction. W04 sample was put through more rigorous selection criteria The latest spectral type secondary we observe is M9/L0. It is thantheWDMsamplebecausetheisolatedMstarspectra,un- likely that later L dwarf companions are hidden by the glare contaminatedbythepresenceofbinarycompanions,weregen- ofthewhitedwarfprimaryandmayhavebeenoverlookedas erallyofhigherqualitythanourseparatedMstarspectra,andthe binaries, although the active ones could be identified by their sheer number of M stars in the SDSS required a method other hydrogenBalmeremissionasdescribedinx3.1(without,how- thanthevisualinspectionofthousandsofMstarspectra.With ever,revealingtheirspectraltype).Whitedwarfswithinactive, thesmallernumberofMstarsintheWDMsample,eachspectrum very low mass secondaries may well be present in the large SDSSwhitedwarfcatalogsofK04andEisensteinetal.(2005), justashigh-massand/orverycoolwhitedwarfprimariesmay 15 ThereismoredetailinvolvedinthedeterminationofactivityinMstars be present in the large SDSS M dwarf catalog of W04 (see thanispresentedhere.WereferthereadertoW04foramorerigorousdiscussion Fig.10,bottom).Infraredphotometryandspectroscopywould ofourMdwarfactivityclassificationmethod.