DTIC ADA464001: A Catalog of Spectroscopically Selected Close Binary Systems from the Sloan Digital Sky Survey Data Release Four

5.4 MB·English

by  Defense Technical Information Center

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.