Table Of Content

DRAFTVERSIONFEBRUARY1,2017 PreprinttypesetusingLATEXstyleemulateapjv.5/2/11 THEMYSTERIOUSDIMMINGSOFTHETTAURISTARV1334TAU JOSEPHE.RODRIGUEZ1,GEORGEZHOU1,PHILLIPA.CARGILE1,DANIELJ.STEVENS2,HUGHP.OSBORN3,BENJAMINJ. SHAPPEE4,5,PHILLIPA.REED6,MICHAELB.LUND7,HOWARDM.RELLES1,DAVIDW.LATHAM1,JASONEASTMAN1,KEIVAN G.STASSUN7,8,ALLYSONBIERYLA1,GILBERTA.ESQUERDO1,PERRYBERLIND1,MICHAELL.CALKINS1,ANDREW VANDERBURG1,ERICGAIDOS9,10,MEGANANSDELL11,ROBERTJ.SIVERD12,THOMASG.BEATTY13,14,CHRISTOPHERS. KOCHANEK2,15,JOSHUAPEPPER16,B.SCOTTGAUDI2,RICHARDG.WEST3,DONPOLLACCO3,DAVIDJAMES17,RUDOLFB. 7 KUHN18,KRZYSZTOFZ.STANEK2,15,THOMASW.-S.HOLOIEN2,15,19,JOSEL.PRIETO20,21,SAMSONA.JOHNSON2,ANTHONY 1 SERGI22,NATEMCCRADY22,JOHNA.JOHNSON1,JASONT.WRIGHT13,14,ROBERTA.WITTENMYER23,24,JONATHAN 0 HORNER23,24 2 1Harvard-SmithsonianCenterforAstrophysics,60GardenSt,Cambridge,MA02138,USA 2DepartmentofAstronomy,TheOhioStateUniversity,Columbus,OH43210,USA n 3DepartmentofPhysics,UniversityofWarwick,GibbetHillRoad,Coventry,CV47AL,UK a 4CarnegieObservatories,813SantaBarbaraStreet,Pasadena,CA91101,USA J 5Hubble,Carnegie-PrincetonFellow 0 6DepartmentofPhysicalSciences,KutztownUniversity,Kutztown,PA19530,USA 3 7DepartmentofPhysicsandAstronomy,VanderbiltUniversity,6301StevensonCenter,Nashville,TN37235,USA 8DepartmentofPhysics,FiskUniversity,100017thAvenueNorth,Nashville,TN37208,USA ] 9DepartmentofGeologyandGeophysics,UniversityofHawai‘iatMnoa,Honolulu,HI96822 R 10CenterforSpaceandHabitability,UniversityofBern,CH-3201Bern,Switzerland S 11InstituteforAstronomy,UniversityofHawai‘iatManoa,Honolulu,HI96822,USA h. 12LasCumbresObservatoryGlobalTelescopeNetwork,6740CortonaDr.,Suite102,SantaBarbara,CA93117,USA p 13DepartmentofAstronomy&Astrophysics,ThePennsylvaniaStateUniversity,525DaveyLab,UniversityPark,PA16802 - 14CenterforExoplanetsandHabitableWorlds,ThePennsylvaniaStateUniversity,525DaveyLab,UniversityPark,PA16802 o 15CenterforCosmologyandAstroParticlePhysics(CCAPP),TheOhioStateUniversity,191W.WoodruffAve.,Columbus,OH43210,USA r 16DepartmentofPhysics,LehighUniversity,16MemorialDriveEast,Bethlehem,PA18015,USA t s 17AstronomyDepartment,UniversityofWashington,Box351580,Seattle,WA98195,USA a 18SouthAfricanAstronomicalObservatory,POBox9,Observatory7935,SouthAfrica [ 19USDepartmentofEnergyComputationalScienceGraduateFellow 2 20NucleodeAstronomadelaFacultaddeIngeniera,UniversidadDiegoPortales,Av.Ejercito441,Santiago,Chile v 21MillenniumInstituteofAstrophysics,Santiago,Chile 4 22DepartmentofPhysicsandAstronomy,UniversityofMontana,Missoula,MT59812,USA 4 23ComputationalEngineeringandScienceResearchCentre,UniversityofSouthernQueensland,Toowoomba,QLD4350,Australiaand 0 24SchoolofPhysicsandAustralianCentreforAstrobiology,UniversityofNewSouthWales,Sydney2052,Australia 3 DraftversionFebruary1,2017 0 . ABSTRACT 1 We presentthe discoveryof two extended ∼0.12magdimming events of the weak-lined T-Tauri 0 starV1334. The startof the firsteventwasmissed butcame toanend inlate 2003,andthe second 7 1 beganin February 2009, and continues asof November 2016. Since the egressof the currentevent : hasnotyetbeenobserved, itsuggestsaperiodof >13yearsifthiseventisperiodic. Spectroscopic v observations suggest the presence of a small inner disk, although the spectral energy distribution i X shows no infrared excess. We explore the possibility that the dimming events are caused by an orbitingbody(e.g. adiskwarpordusttrap),enhanceddiskwinds,hydrodynamicalfluctuationsof r a theinnerdisk,orasignificantincreaseinthemagneticfieldfluxatthesurfaceofthestar.Wealsofind a∼0.32dayperiodicphotometric signalthatpersiststhroughoutthe2009dimmingwhichappears to not be due to ellipsoidal variations from a close stellar companion. High precision photometric observations of V1334Tau during K2 campaign 13, combined with simultaneous photometric and spectroscopicobservationsfromtheground,willprovidecrucialinformationaboutthephotometric variabilityanditsorigin. Subjectheadings:circumstellarmatter,protoplanetarydisks,stars: individual: V1334Tau,stars: pre- mainsequence,stars: variables:TTauri 1. INTRODUCTION and in many cases stellar companions. Many of these characteristics can manifest themselves through disk The circumstellar environments of young stellar ob- substructures,gradients,andotherdiskpropertiesthat jects(YSOs)arecomplexbecausealargenumberofpro- could revealthe mechanisms that influence planet for- cesses drive their evolution. The observed properties mation. Although high spatial resolution imaging and of YSOs involve accretion onto the star, dispersion by interferometry can now resolve the circumstellar envi- stellar winds and radiation, magnetic fields, outflows, ronment of some young stars, we are not yet able to 2 Rodriguezetal. probe the innermost region for most stars (<10 AU) (WTTS)V1334Tau. Herewe find two dimming events wherealargefractionoftheexoplanetsreside. of V1334 Tau, one that ended in late 2003 and another First classified by Herbstetal. (1994), YSOs are thatbeganin2009andextendstothepresentdate. The known to vary in the optical as a result of material ac- known characteristics of the V1334 Tau system are de- cretingontothestellarsurfaceand/orcircumstellarex- scribed in §2. The photometric observations are pre- tinction. This variability can vary in duration, depth, sented in§3. We presentour estimate of the stellar pa- and periodicity. Some YSOs showing non-periodic, rametersofV1334TauAin§4. Wedescribethespectro- large (>1 mag) and long (months to years) dimmings scopicanalysisandthelong-termphotometricvariabil- and are referred to as UX Orionis stars (UXors). Pro- ity in §5, discuss possible dimming mechanisms in §6, posedcausesareover-denseregionsinthenearlyedge- andsummarizeourresultsin§7. oncircumstellardisksurroundingthehoststar(Wenzel 1969; Grinin 1988; Voshchinnikov 1989; Grininetal. 2. THEV1334TAUSYSTEM 1998;Gradyetal.2000),hydrodynamicalfluctuationsof V1334 Tau was previously identified as a Weak- the inner disk (Dullemondetal. 2003), or strong disk lined T Tauri Star (WTTS) with a K1 spectral type winds(Petrov&Kozack2007). by Wichmannetal. (1996) using observations from the Shorter duration (days to weeks) photometric vari- ROSATAll-Sky-Survey(Zimmermannetal.1993)com- ability of YSOs has been seen in both the optical and bined with optical spectra. More recently, V1334 Tau infraredbutthecauseisunclear. Thisvariabilitycould was re-classified as a G2 star (Nguyenetal. 2012). be caused by accretion of circumstellar material onto Wichmannetal. (2000) determined that V1334 Tau has the stellar surface causing hot spots, circumstellar ex- a radius of 2.73 R⊙, a mass of 2.05 M⊙, and an age of tinction, or orbiting dust clumps (Herbstetal. 1994; 3.83×106yr. V1334Tauhasapropermotionofµ =8.7± α Stassun&Wood 1999; Bouvieretal. 2007). These stars 0.8andµ =−24.7±0.7masyr−1whichisinagreement aretypicallyreferredtoas“dippers”,andalthoughthe δ with the known motion of the Taurus-Aurigaeassocia- variabilityisshorterinduration,thedimmingcanhave tion(µ =7.2andµ =-20.9,Bertout: &Genova2006).Us- a depth of up to ∼1 mag. Dippers have a lower am- α δ ingspeckle observations, Kohler&Leinert(1998)iden- plitude in the infrared than the optical, supporting the tifiedaclosecompaniontoV1334Tau(1′.′631,∆K=3.42). interpretationthattheyarecausedbycircumstellardust A near-IR high-spatial-resolution survey for low-mass obscuration(Codyetal.2014). companions in Taurus by (Daemgenetal. 2015) found Auniquewaytostudyprotoplanetaryformationand ′′ two additional nearby companion candidates at 0.106 evolution is to observe YSOs that exhibit significant large dimming events (>10% depth and months to witha∆K=1.68±0.05and6′.′969witha∆K=12.5±0.3. yearsin duration) causedbya portion of the star’scir- 3. OBSERVATIONS cumstellardisk. Whiletheseeventsarequiterare,they Over the past decade multiple ground-based pho- provide a powerful tool for studying the circumstellar tometric surveys have observed V1334 Tau. None of environment of YSOs. Using high cadence photomet- the photometric observations resolve the fainter com- ric observations from the Kilodegree Extremely Little ponents in the system. All the follow-up photometry Telescope(KELT,Pepperetal.2007,2012),wehavebeen showninFigure1isavailableinmachine-readableform conducting the Disk Eclipse Searchwith KELT (DESK) intheonlinejournal. survey to identify and characterize these rare systems (Rodriguezetal. 2016a). The DESKsurvey has discov- 3.1. KELT ered and analyzed previously unknown disk eclipsing events around the stars RW Aurigae (Rodriguezetal. Withtheprimarygoalofdiscoveringtransitingplan- 2013, 2016d), V409 Tau and AA Tau (Rodriguezetal. ets around bright host stars (8 < V < 11), the KELT 2015),TYC2505-672-1(Rodriguezetal.2016b),andDM project is a ground-based photometric survey cover- Ori(Rodriguezetal.2016c). ing >70% of the entire sky on a 10-20 minute cadence Interestingly, not all of these discoveries are YSOs. (Pepperetal. 2007, 2012). The project consists of two TYC2505-672-1isanevolvedMgiantbeingeclipsedev- telescopes,KELT-NorthinSonoita,AZandKELT-South ery∼69yearsbyacompanionwithadisk. Thissystem in Sutherland, South Africa. Each telescope has a isverysimilartothearchetype,ǫAurigae,anF-giantbe- Mamiya 645-serieswide-angle lens with a 42mm aper- ingeclipsedevery∼27yearsbyasmallstarembedded ture and a 80mm focal length (f/1.9), and observes in inacircumstellardisk(Carrolletal.1991).Interferomet- a broad R-band filter. The telescopes have a 26◦ × 26◦ ′′ ric observations by the Georgia State Universitys Cen- field-of-view (FOV), and a 23 pixel scale. V1334 Tau ter for High Angular Resolution Astronomy (CHARA) islocatedinKELT-Northfield03(fieldcenteratα =3h ◦ ′ ′′ duringthelasteclipseofǫAurconfirmedthisenterpre- 58m12s,δ=31 39 56.16 ). TheKELT-Northtelescope tationbyimagingthecompanionanddiskcrossingthe observed V1334 Tau from UT 2006 October 26 to UT star (Kloppenborgetal. 2010). These evolved systems 2013 March 13, obtaining 9186 observations after pro- still have remnant circumstellar material in their sys- cessing. ForadetaileddescriptionoftheKELTdataac- tems.Sincethesediscoveriesareallatdifferentagesand quisitionandreductionprocess,seeSiverdetal.(2012). stagesofdiskevolution,eachsystemprovidesinforma- Themedianper-pointerroris0.005mag. tion about differentstagesof protoplanetaryformation 3.2. All-SkyAutomatedSurveyforSuperNovae(ASAS-SN) and evolution which may in turn provide insight into thelargediversityofexoplanetarysystems. Designedtodiscovernearbysupernovae,theAll-Sky In this paper, we presentnewphotometric and spec- AutomatedSurveyforSuperNovae(ASAS-SN)is pho- troscopic observations for the Weak-lined T Tauri Star tometrically monitoring the observable sky every two V1334Tau 3 FIG.1.—TheKELT(black),ASAS(red),WASP(blue),ASAS-SN(purple),KUO(yellow),andMINERVA(green)photometricobservationsof V1334Taufrom2002untillate2016. OnlytheV-bandobservationsfromKUOandMINERVAareshown. Theredverticallinesrepresentthe estimatedendofthepre-2004eventsandthestartofthe2009event. Thegreyshadedregionrepresentstheestimatedegress/ingressforeach eventdescribedin§5.Theintercalibrationofthedatasourcesisdescribedin§5. days down to V ∼ 17 (Shappeeetal. 2014). The sur- second SuperWASP data set in 1-hour bins and calcu- veyhastwosites,CerroTololoInterAmericanObserva- latedthestandarddeviationofeachbin. Alldatapoints tory (CTIO) in Chile and Mount Haleakala in Hawaii. withinbinswhichhadastandarddeviationgreaterthan Each site is hosted by the Las Cumbres Observatory 2% were then discarded. After processing and clean- Global Telescope Network (Brownetal. 2013) and has ing,14339observationsremainedandwereusedinthis twounitswithfour14cmNikontelephotolensesanda work. The median per-point error is 0.009mag for the 2k×2kthinnedCCD,witha4.5×4.5degreeFOVanda publicobservationsand0.003magfortherecentobser- 7′.′8pixelscale. V1334Tauwasobserved532timesfrom vations. UT2012January20untilUT2016October14. Theme- dianper-pointerroris0.004mag. 3.4. All-SkyAutomatedSurvey(ASAS) Designed to discover and catalog variable stars over 3.3. SuperWASP the entire sky, the All-Sky Automated Survey (ASAS) The Wide Angle Search for Planets (SuperWASP) is observes all stars brighter than V=14. For a detailed a high-cadence photometric survey designed to detect description of the observing strategy and data reduc- transiting exoplanets. Observing in a broad filter (cen- tion, see Pojmanski (1997). ASAS uses two locations, tered at 550 nm), the survey has two observing loca- Las Campanas, Chile and Haleakala, Maui. Each ob- tions, one atthe Roque de los MuchachosObservatory serving location observed in the V and I bands simul- on La Palma (WASP-North) and the other located at taneouslyusing twowide-fieldMinolta 200/2.8APOG the South African Astronomical Observatory (WASP- telephotolenseswitha2K×2KApogeeCCD.Eachtele- South).With8camerasateachlocation,eachusinga200 scopehasa8.8◦ ×8.8◦FOV.ASASobservedV1334Tau mmf/1.8lensanda2048×2048pixelCCD,SuperWASP intheV-bandfromUT2002December13untilUT2009 isabletomonitorlargeportionsoftheskyataveryhigh February 25. 93 observations were obtained over this cadence(minutes). Eachcamerahasa7.8◦ ×7.8◦ FOV, period. Themedianper-pointerroris0.034mag. ′′ anda13.7pixelscale. V1334Tauwasobservedintwo separate seasons in the publicly released observations: 3.5. MINERVA UT 2004 August 15 to UT 2004 September 30 and UT 2006September11untilUT2007February151. Allob- The MINERVA Project has four 0.7m PlaneWave CDK-700telescopeslocatedonMt. Hopkins,AZ,atthe servations from WASP prior to 2006 are unfiltered and FredL.WhippleObservatory,capableofbothmillimag after2006,theobservationsareinabroadV-bandfilter. photometryandprecisionradialvelocitymeasurements A 3σ clipping around the median was done to remove (Swiftetal.2015). Onthe nightof UT2016October21, large outliers. V1334 Tau was also observed from UT twotelescopessimultaneously observedV1334Tauus- 2009 February 08 until UT 2012 January 29. Since the datareductionofthe public SuperWASPdatawasper- ing an Andor iKON-L 2048×2048 detector with a 20.′9 formeddifferentlythanthemorerecentWASPobserva- × 20.′9 FOV and a plate scale of 0′.′6 pixel−1. One tele- tions, the two data sets were kept separate throughout scope observed in the V-band with 200 exposures and theanalysisdescribedhere.Additionally,webinnedthe achieved a per-point RMS scatter of 0.0009 mag. The other observed in the B band with 140 exposures and 1http://exoplanetarchive.ipac.caltech.edu/ achievedaper-pointRMSscatterof0.0015mag. 4 Rodriguezetal. TABLE1 STELLARPROPERTIESANDPHOTOMETRICMEASURMENTSOFV1334TAUOBTAINEDFROMTHELITERATURE. Parameter Description Value Source Reference(s) Names V1334Tau HD290380 TYC1839-643-1 2MASSJ04445445+2717454 αJ2000 RightAscension(RA) 04:44:54.454 Tycho-2 Høgetal.(2000) δJ2000 Declination(Dec) +27:17:45.23 Tycho-2 Høgetal.(2000) BT TychoBTmagnitude 10.909±0.061 Tycho-2 Høgetal.(2000) VT TychoVTmagnitude 9.895±0.038 Tycho-2 Høgetal.(2000) FUV FarUVmagnitudes 20.05 GALEX Go´mezdeCastroetal.(2015) NUV NearUVmagnitudes 15.63 GALEX Go´mezdeCastroetal.(2015) V JohnsonV 9.612±0.04 APASS Hendenetal.(2016) B JohnsonB 10.511±0.063 APASS Hendenetal.(2016) g′ Sloang’ 10.312±0.045 APASS Hendenetal.(2016) r′ Sloanr’ 9.361±0.06 APASS Hendenetal.(2016) i′ Sloani’ 8.923±0.05 APASS Hendenetal.(2016) J 2MASSmagnitude 7.734±0.02 2MASS Cutrietal.(2003) H 2MASSmagnitude 7.281±0.04 2MASS Cutrietal.(2003) Ks 2MASSmagnitude 7.154±0.02 2MASS Cutrietal.(2003) WISE1 WISEpassband 7.017±0.033 WISE Cutrietal.(2012) WISE2 WISEpassband 7.060±0.018 WISE Cutrietal.(2012) WISE3 WISEpassband 7.048±0.018 WISE Cutrietal.(2012) WISE4 WISEpassband 7.058±0.108 WISE Cutrietal.(2012) µα ProperMotioninRA(masyr−1) 8.7±0.8 NOMAD Zachariasetal.(2004) µδ ProperMotioninDEC(masyr−1) −24.7±0.7 NOMAD Zachariasetal.(2004) RV....... Systemicradial ............... ∼19 thiswork velocity(kms−1) U∗....... Spacemotion(kms−1).......... −12.1±0.6 thiswork V........ Spacemotion(kms−1).......... 24.3±1.6 thiswork W........ Spacemotion(kms−1).......... 16.2±2.4 thiswork NOTES Thephotometryshownisusedin§4todeterminethestellarparameters.U∗ispositiveinthedirectionoftheGalacticCenter.UVWanalysis suggestsa98.4%chanceofV1334TaubeinginthethindiskaccordingtotheclassificationsystemofBensbyetal.(2003).TheUVWkinematics usedthepeculiarvelocityoftheSunwithrespectlocalstandardrestfromCos¸kunogˇluetal.(2011).TheUVWanalysisanddistancefrom§4are consistentwiththisbeingamemberoftheTaurus-Aurigaeassociation. 3.6. KutztownUniversityObservatory Whipple Observatory, on Mount Hopkins, Arizona, USA from UT 2016 October – December. We obtained Usingthe0.61mRitchey-Chre´tienopticaltelescopeat nineteen10minute exposuresover thespanofthe two the Kutztown University Observatory (KUO) in Kutz- months, with fourteen observations occurring on UT town, Pennsylvania, V1334 Tau was observed in the B 2016 December 09 and 10. The spectra were obtained andV-bandson UT 2016October 26, UT2016Novem- ber 05 and 07. The observing setup uses a 3072×2048 at a resolution of λ/∆λ ≡ R = 44000 over the wave- CCD that has a 19.′5×13.′0 FOV with a 0′.′38 pixel−1. lteenngttwhirthanage∼o5f503090K0d−w9a1r0f0aA˚n.dTohuerspesetcitmraataerefrocomns§is4-, Standard aperture photometry was used, and the in- rotationallybroadenedby77.6,kms−1. Weestimatethe strumental flux was color-corrected using several Lan- absolute radial velocity of the V1334 Tau system to be doltstandardfields. Intotal,KUOobtained160B-band and 186 V-band observations. The median per point a ∼19 kms−1whichisinagreementforthe knownabso- per-point RMS is 0.0007mag in the B-band and 0.0006 luted radial velocity of the Taurus-Auriga association V-band. (15 kms−1, Bertout: &Genova 2006). The TRES observations show no large (>1 kms−1) radial velocity 3.7. SpectroscopicFollow-up changes. We obtained spectroscopic observations of V1334 TauwiththeTillinghastReflector EchelleSpectrograph (TRES) on the 1.5m telescope at the Fred Lawrence 4. DETERMINATIONOFSTELLARPARAMETERS V1334Tau 5 TABLE2 −15 DETERMINEDSTELLARPARAMETERSFORV1334TAUAAND68% CONFIDENCEINTERVALVALUES −16 Parameter Description DeterminedValue Age StellarAge 1+1Myr −1 −17 M⋆ StellarMass 1.674+−00..222799 M⊙ R⋆ StellarRadius 2.619+−00..539074 R⊙ 21Å ]−−−18 logT(L⋆) LEoffgecSttievlelaTreLmupmeirnaotusirtey 0.655081+−5+00..91230134KL⊙ m eff −154 1 c− log(g) SurfaceGravity 3.767+−00..101845cgs g(F) [erg secλ−−2109 [[FFeDe//iHsHta]]sniuncrietfiaacle M..e.St.au.lr.lfi.ac.ic.tey..Ma.t.e.fto.a.rl.lmi.c.at.yti.o.n. −−1003..10217+−6731+−+−240000P....0101c87863114 lo Av Extinction 0.438+−00..019620mag −21 Weuse MINESweepertoinferthephysicalproperties −22 (e.g.,mass,radius,age,etc.) ofV1334Tau,aswellasits distanceandextinction(A ). Wemodeledalltheavail- V ablebroadbandopticalandIRphotometry(seeTable1) −23 103 104 105 excepttheGALEXUVandAPASSphotometry. Weare hesitant to include the GALEX data due to the likely Wavelength [Å] presence of chromospheric emission in such a young FIG.2.— Spectral energy distribution of V1334 Tau along with rapidlyrotatingstar.TheAPASSdatawerenotincluded a model using the most probable stellar parameters from our due to our past experience with unaccounted for zero- MINESweeperanalysis. ThelightbluepointsshowtheTycho-2BT, VT,2MASSJ,H,Ks,andWISE1–4photometricobservationsincluded point offsets seen in this survey’s photometry (Cargile inthefit.ThedarkbluepointsaretheGALEXandAPASSphotometry et al. in prep). We applied a Gaussian distance prior thatwerenotusedinthefit. TheUVexcessislikelyduetochromo- centered at 140 pc with a FWHM of 40pc (Torresetal. sphericactivity. 2007,2009),andauniformageprior(0−50Myr). InordertodeterminethestellarparametersofV1334 Thespectralenergydistribution(SED)forV1334Tau Tau,weuseMINESweeper,anewlydevelopedBayesian doesnothaveaninfrared(IR)excessindicativeofacir- approachto determining stellar parameters. A full de- cumstellar disk (SeeFigure 2), butthere is anexcessin scription of MINESweeper is given in Cargile et al. (in theultra-violet(UV)whencomparedtotheMISTstellar prep);hereweprovideabriefsummaryofthemethod. SED model. In Figure 3, we show the posterior prob- MINESweeper uses nested importance sampling to de- abilitydistributions for the age, mass, radius, distance, termine posterior probability distributions for physi- andAV.Ourinferreddistanceandageagreeswithstan- cal parameters inferred from stellar evolution models. dard estimates for solar-type stars in the Taurus star- MINESweeper uses a modified version of the nestle.py forming region, ∼130±25 pc (Torresetal. 2007, 2009) code2toperformmulti-nestedellipsoidsamplingbased and ∼1 Myr (Bricenõetal. 2002; Luhmanetal. 2003) – suggesting that V1334 Tau appears to be a bona fide onthealgorithmdescribedinFerozetal.(2009). Multi- memberofthisstellarassociation. Ourmodelssuggest nested sampling is uniquely suited to the problem thatthemassofV1334Tauislowerthanthatproposed of modeling parameters from stellar evolution models byWichmannetal.(2000)butwithin1σ(SeeTable2for due to its ability to efficiently sample the multi-modal ourdeterminedstellarparameters). likelihood surfaces, frequently found when modeling stars with stellar isochrones. MINESweeper uses the 5. RESULTS most recent release of the MIST stellar evolution models (Choietal. 2016), and an optimized interpolation In this section, we examine the available photome- schema based on the recommendations of Dotteretal. try. We explore different interpretations for the long- (2008). MINESweeper can use both photometric and durationdimmingeventsobservedin§7.Allphotomet- spectroscopicmeasurementsaswellasawiderangeof ricobservationsdisplayedinFigure1areintheV-band prior probability distributions. The MINESweeper in- (ASAS, ASAS-SN, KUO, and MINERVA) or a broader ferenceresultsinposterior probabilitydistributionsfor filter (KELT and WASP). To place all photometric data thefundamentalMISTmodelparameters: equalevolu- onthesamescale,weusetheKUOobservationsasthe tionarypoints(EEP),inputstellarmetallicity,andstellar photometricstandardandapplyaverticaloffsettoeach age. Usingthese posterior distributions, MINESweeper data set to align them where they overlap. We do not thengeneratesposteriordistributionsforanyotherstel- otherwiecorrectforthefilterdifferences. lar parameters predicted by the MIST models such as the mass, radius, T , luminosity, and surface abun- 5.1. Out-of-DimmingVariability eff dances. We searched for periodic variability in the high- cadence KELT data prior to and during the 2009 dim- 2http://kbarbary.github.io/nestle/ mingusingtheLomb-Scargle(LS)periodicitysearchal- 6 Rodriguezetal. 3.0 ]⊙2.5 M s [2.0 s a M1.5 1.0 4.0 ]⊙3.5 R3.0 s [2.5 u di2.0 a R1.5 1.0 300 250 c] Dist [p125000 100 1.2 1.0 0.8 AV0.6 0.4 0.2 0.0 0 2 4 6 8 101.0 1.5 2.0 2.5 3.01.0 1.5 2.0 2.5 3.0 3.5 4.0 100 150 200 250 3000.0 0.2 0.4 0.6 0.8 1.0 1.2 Age [Myr] Mass [M ] Radius [R ] Dist [pc] AV ⊙ ⊙ FIG.3.—PosteriorprobabilitydistributionsofthestellarparametersforV1334TaubasedonourMINESweeperanalysis.TheparametersinTable 2arethemaximumposteriorprobabilitiesoftheseposteriorsalongwiththeir68%confidenceintervals. gorithm (Lomb 1976; Scargle 1982) in the VARTOOLS orbital period of some circumstellar material that peri- analysis package (Hartman 2012). For the each season odicallyobscuresthestaratasemimajoraxisof∼2.3R⋆. of KELT data, we find a 0.3214day period with an av- However, the observed periodicity is persistent over erage peak-to-peak amplitude of ∼2% (see Figure 4). the entire KELT data set, making it unlikely to be re- This periodic signal is clearly changing in amplitude lated to accretion processes. Additionally, if there is a and phase over the 9 years of observations. We find significantamountoforbitingdiskmaterialrightatthe the same ∼0.32 day periodicity in every season except surfaceofthestar,wewouldexpecttoseeanIRexcess the2008-2009observingseasonwherethesourcefaded. intheSEDanalysis. ThefourteenspectraofV1334Tau Additionally, the targeted observations by MINERVA on UT 2016 December 09 and 10 show no significant and KUO are consistent with this periodicity. There large radial velocity variability ruling out the possibil- is no color evolution over the short duration of these ityof aclosestellarbinary. The orbitalsemi-amplitude targeted observations. During the 2009 dimming, the from the TRES observations is 338±56 ms−1, when a ∼0.32dayperiodicityisrecoveredbutatalowersignal- photometricephemeriswithaperiod0.3124daysisim- to-noise ratio. The periodicity is likely caused by pro- posed. Thevelocityresidualsfromacircularorbithave cesses near the star. If the 0.32 day periodicity is the an RMS of 433 m/s. The RMS of all the RV measure- rotation period of the star, then this would imply a ro- mentsis1.3 kms−1. Usingourdeterminedstellarmass tational velocity of ∼ 347±57 kms−1for a stellar ra- of 1.67 M⊙and a period of 0.3214 days, we get an up- odfiu0s.3o9f±2.06.1193+−d00..a5390y74sR. ⊙G,ivceonmopuarreudntcoeratabinretyakinupthpeesritoeld- pmesr−m1)asosrl<im6i.t2oMfaJc(RomMpSa=ni1o.n3tkombes−<12). MThJe(RuMpcSo=m4in3g3 larradius,V1334TauAcouldberotatingnearbreakup K2campaign13observationsofV1334Taushouldpro- velocity. We explore the possibility that V1334Tau is a videadditionalinformationonthenatureofthisshort- rapidlyrotatingstarclosetoapole-onorientationin§6. periodvariability. Another possibility is that the ∼0.32 day period is the V1334Tau 7 13.30 13.35 13.40 K 13.45 R 13.50 13.55 13.60 4000 4500 5000 5500 6000 6500 7000 JD-2450000 [d] 1.06 0.30 1.04 0.25 1.02 0.20 0.15 1.00 0.10 0.98 0.05 0.96 0.00 x)1.06 0.30 Flu1.04 0.25 ( P0.20 an1.02 LS0.15 edi1.00 0.10 m0.98 0.05 x/0.96 u 0.00 Fl1.06 0.30 1.04 0.25 1.02 0.20 1.00 0.15 0.10 0.98 0.05 0.96 0.00 1.5 2.0 2.5 3.0 1.5 2.0 2.5 3.0 1.5 2.0 2.5 3.0 0.0 0.2 0.4 0.6 0.8 0.0 0.2 0.4 0.6 0.8 0.0 0.2 0.4 0.6 0.8 Frequency (1/day) Phased @ Period = 0.3214 d FIG.4.—(Top)The9yearlightcurveofV1334Taucolorcodedbyobservingseason. (Bottom-left)TheLSperiodogramsforeachKELTseason. Theverticallinescorrespondtothe0.32dayperiodestimate(black)andtwicethatperiod(Red).(Bottom-right)EachseasonofKELTdataphase- foldedtothe0.32dayperiod.Notethatthephaseandamplitudeofthe0.32daysignalischangingoverthe9observingseasonsfromKELT. 5.2. Pre-2004Dimming takes ∼273 days and that the event has a maximum depth of ∼0.12 mag (V ∼ 9.71, the maximum depth In UT 2002 December V1334 Tau was in its “minimum” state, at a V ∼ 9.7 mag. The system took ∼280 occurredduring the 2012and 2016observing seasons). days to return to its quiescent brightness of ∼9.6 mag. This ingress timescale is very similar to the estimated timescale to brighten in 2002-2003. The dimming lasts Unfortunately, there are no available photometric ob- through the end of the data analyzed here on UT 2016 servationsof V1334TaupriortoUT2002December, so November07,leadingtoalower limitonthe dimming thereisnoconstraintonthedimmingduration.Weonly durationof>2802days. Ifweassumethatthetwodim- knowthatthedimmingmustbelongerthanthetimebe- mingeventsaresimilar,thetimebetweeneventsis>13 tweenthestartoftheASASobservationsandtheendof theobservedbrightening-i.e. greaterthan∼320days. years. Multi-band observations from MINERVA and KUO show no B−V color trend over the course of an 5.3. 2009Dimming ∼6hourobservingnight. With the combined high cadence observations from During late February of 2009, the V1334 Tau system KELTandWASPduringthe2009dimming,weareable began to fade from its median magnitude of V ∼ 9.59 to study the photometric variability during the event. mag. We estimate that the ingress of the dimming 8 Rodriguezetal. If this is indeed an occultation, the additional bright- widths of 1.5A˚ and 2.6A˚ respectively, perhaps indica- ening and dimming events seen during the large dim- tiveofthestrongchromosphericactivityexpectedforT ming(onasimilartimescaleastheinitialdimming)sug- Tauristars. Lithiumabsorptionat6708A˚ isalsopresent, geststhatwe areobserving sub-structureinthe occult- with an equivalent width of 0.25A˚, consistent with the ing body. After the ingress, the system stays at a con- absenceofdepletionexpectedforasolar-typestarwith stant brightness of ∼9.66 mag through the remainder anageof2Myr(Soderblometal.1993). ThesodiumNa of the 2009-2010 observing season and the first half of IDfeaturehasadepthandwidthconsistentwithastel- the 2010-2011season. At JD ∼ 2455502,V1334Tau be- larabsorptionline, alongwith asuperimposed narrow gins to dim again, down to brightness of ∼9.69 mag. featureduetotheinterstellarmedium. UnlikeRWAur At the start of the 2011-2012 observing season, V1334 (Facchinietal. 2016), we do not see a time varying P Tauisnearitsminimumbrightnessof∼9.71mag. Since CygniprofileintheNalines, whichwould suggestthe theendofthisadditionaldimmingoccurredduringthe existence of a strong disk wind. However, we do not 2011 seasonal observing gap, we can only place a up- havea spectroscopic observationof V1334Tauprior to perlimitof∼325daysonitsduration. AfterV1334Tau thelargedimming,limitingourinterpretation. brightensto∼9.65in2012,itremainsatadepthof∼9.69 magforthefollowingseason. Followingthe2013/2014 6. INTERPRETATIONANDDISCUSSION season,V1334Taubrightensoverthecourseoftheentire season from ∼9.69 mag to ∼9.64 mag. Following this With only two incompletely sampled dimming events, we are unable to determine whether this peakinbrightness,thefollowingtwoseasonsofASAS- phenomenon is periodic in nature. However, the SN, KUO, and MINERVA data suggest that V1334Tau fadedbacktothemaximumdepthof∼9.71mag. Inter- egress/ingress timescales from each event are consistent. In similar systems with dimming events, such as estingly, the timescales of these variation aresimilar to V409TauandDMOri,thedimmingsarebelievedtobe the estimated ingress timescale in 2009. However, our periodic (Rodriguezetal. 2015, 2016c). In this section ability to draw any conclusions is limited by the sea- we explore the possibility that the dimmings of V1334 sonal observing gaps. We discuss the possibility that Tauis caused by an orbiting body (e.g. a disk warp or the observed structure of the dimming may be caused dusttrap),enhanceddiskwinds, hydrodynamicalfluc- byachangeintheopacityoftheoccultingfeaturein§7. tuationsoftheinnerdisk,orasignificantincreaseinthe magneticfieldfluxatthesurfaceofthestar. 5.4. SpectroscopicAnalysis To check for photospheric features on the stellar 6.1. OccultationbyOrbitingBody disk,wederivedlinebroadeningprofilesusingaleast- To determine the key propertiesof the occulting fea- squares deconvolution (LSD) of the TRES spectra (fol- ture, we model the observed dimmings as an occulta- lowing Zhouetal. 2016). Indentations and protrusions tion of the host star by a large body in Keplerian mo- inthebroadeningprofilemapthepresenceofspots,fac- tion. The leading edge of the occulting body can be ulae,andpulsationonthestellarsurface. Thebroaden- perpendicular to its direction of motion (θ = 90, sharp ingprofilesshowlarge-scalevariationsoverthecourse edge)orcanbeinclined(“wedge-shaped”,seeFigure9 of hours. Figure 5 shows the time series of the spec- fromRodriguezetal.2015). Theestimatedingresstime tra gathered over 2016December 09 and 10, phased to ofthe2009dimming,wherewehavethebestphotomet- the0.3214dayphotometricmodulationperiod(seeSec- riccoverage, is ∼273days. Theingresstimescale gives tion 5.1). We find that the broadening profiles do not usanestimateoftheoccultingbody’stransverseveloc- phasewellwiththephotometricperiod,suggestingthat thereisnopersistentspotcrossingthestar,andthatthe ity, VT ∼ 2R⋆/(Tingresssinθ). For R⋆ = 2.62R⊙, we get variability is possibly due to non-radial pulsations on a minimum velocity of ∼155.1 ms−1. The full dura- thestar. Inaddition,wefindnoevidenceoflineblend- tionof the2009dimming is >2803days. Usingour es- ingindicativeofabinarystellarcompanionorbitingthe timatedtransversevelocityandthedurationtimescale, primarystar. weestimatethattheocculting bodymustbe >0.25AU Wealsofindadouble-peakedHαemissionline,which inwidth. IfweassumethattheoccultingbodyisinKe- maybeanindicationofanaccretiondisk(SeeFigure5). plerian motion, the ingress timescale also provides an TheHαfeaturehasanequivalentwidthof∼5A˚,which estimate of the orbital semi-major axis (see Equation 1 could be taken to mean that the star still possesses an fromRodriguezetal. 2016c). However, using these es- inner accretion disk (Mohantyetal. 2005). The rela- timatedparameters,theoccultingbodywouldbeunre- tively high equivalent width also suggests that V1334 alisticallyfarfromthehoststar(at∼62,000AU),ruling Tau should be reclassified as a Classical T-Tauri Star thisinterpretationout. (CTTS)(>3A˚,White&Basri2003).Givenourmassesti- Unfortunately,thereisnomeasurementofthesizeof mates, the peakHα emission correspondstoadistance the circumstellar disk around V1334 Tau A. We know of>0.16AUfromthestar,assumingweareobservinga fromKohler&Leinert(1998)andDaemgenetal.(2015) Keplerianvelocityforapossibleinnerdisk. However,it thatthere aretwo nearby companions to V1334TauA, ′′ ′′ isalsopossiblethattheHαfeaturecouldbefromcooler, V1334 Tau B (1.631) and C (0.106). Therefore, we can neutral HI that is falling onto the star or disk winds. constrain the occulting feature to be within the pro- ′′ NootherBalmerlinesarefoundinemission,nordowe jected separation of the B component, 1.631 (215 AU findothercommonaccretionsignatures,suchastheHe for d = 132 pc). Assuming the occulting feature is at I 5876A˚ and He I 6670A˚ lines in emission. We see the the projected separation of V1334 Tau B, it would re- CaIIHandKlineswithcoreemission, andequivalent quire a transverse velocity of ∼2.6kms−1 and be ∼4.2 V1334Tau 9 FIG.5.—TheHαspectrallineprofilesofV1334Taushowdramaticvariationsoverthecourseofhours.Left:Thedouble-peakedHαlineprofile fortheTRESobservationsonUT2016December09(Red)and10(Blue)phasedtothe0.32dayphotometricmodulationperiod(seeSection5.1). Notethatthelineprofilechangessignificantlyoverthespanoftheobservations. Right:RotationalbroadeningprofilesderivedfromeachTRES spectrum,derivedviaaleast-squaresdeconvolutionprocessandagainphasedtothe0.32dayperiod. AUinwidth. Also,constrainingtheocculterto219AU model, the leading edge of the occulting body would wouldsuggestawedge-angleof<3.4◦. Fromouranaly- have a highly inclined leading edge angle of ∼0.6◦, al- sisoftheTRESspectra,weestimatethatthepeakemis- mostparalleltoitsdirectionofmotion. Thiswouldalso sion of the double-peakedHα emission corresponds to suggestthattheocculterismovingat∼15kms−1andbe adistanceof>0.16AU.Assumingtheoccultingfeature ≥24AUinwidth,whichis>59%oftheentireorbit. is associated with an inner disk and at a semi-major Observations from the Atacama Large Millime- axis of 0.16 AU would imply that it is moving at ∼98 ter/submillimeterArray(ALMA)andtheHubbleSpace kms−1and have a leading edge angle of 0.06◦. In this Telescope(HST)haveshownthattransitionaldiskscan scenario,thedurationoftheoccultationrequiresmulti- have large asymmetries (Pe´rezetal. 2014; Starketal. ple orbits (≥160) of the same material across our line- 2014). Specifically, HST observations of MP Mus have of-sighttocausethedimmingeventthatbeganin2009. shown that the eastern side of its disk is three times Thiscouldbeexplainediftheentireinnerdiskhaspre- brighter then its western side (Schneideretal. 2014). cessedintoourlineofsightbutthisdoesnotseemplau- The young star Oph IRS 48 has a large asymmetric sible. We can also place the occulting body at the dis- dust structure that extends across almost one third of tance of the C component (14AUfor d =132pc). This the entire disk ring (vanderMareletal. 2013). There- would suggest a transverse velocity of ∼10.3kms−1 , fore,ifV1334Tauhasanedge-on,asymmetricdisk,the a ∼16.7AU wide occulter, and aleadingedge angle of large dimming could be caused by a dust trap in the 0.86◦.Thewidthwouldbe∼19%oftheentireorbitat14 diskcrossingourline-of-sightleadingtoaperiodicdim- AU, assuming Keplerian motion. At 14 AU, an orbital ming. FuturemillimetermappingoftheV1334Tausys- periodwouldbe∼40years. Itispossiblethattheclose- temshoulddeterminetheplausibilityofthisinterpreta- in nearby companion is associated with the observed tion. However,nodetectableIRexcessoutto20microns dimmingevents. wouldsuggestaverylowmass,nearlytransparentcir- Fortunately,weobservedtwoseparatedimmings,one cumstellardisk,makingthisinterpretationunlikely. pre-2004 and one beginning in 2009. Since the egress fromthepre-2004dimming andthe ingressof the2009 6.2. InnerDiskObscuration dimming are similar in duration, it is plausible that An alternate explanation for the UXor phenomena we are seeing a periodic occultation with a period of is hydrodynamical fluctuations of the inner disk rim. >13 years. For a stellar mass of 1.67 M⊙, this period Dullemondetal. (2003) determined that the inner disk wouldsuggestthattheoccultingfeatureislocated≥6.6 rimwould have a scaleheight of ∼0.2R and fluctu- rim AU from the host star. Using the same wedge-shaped ateby±∼0.1R . Using0.16AUasanestimateofthe rim 10 Rodriguezetal. innerdiskradius,thiswould suggestamaximumfluc- The complex circumstellar environment of YSOs can tuationoftheinner rimheighttobe ∼0.048AU.Inad- directlyinfluenceplanetaryformationandarchitecture. ditiontothehydrodynamicalfluctuation,theinnerdisk Unfortunately,itisnotclearhoweachprocessorfeature ofaYSOcanbewarpedandperturbedbyamisaligned of a circumstellar disk (accretion, winds, warps, and close companion (Facchinietal. 2013, 2014). Without a companions) affects the planet forming process. From constraintontheinnerdiskinclination,weareunableto studying these processesfor a large sample of YSOs at test the feasibility of this interpretation. However, this different ages, we may gain insight into the role each scenariohasbeenproposedasthecauseoftheobserved featurehasonplanetformation. dimmingeventsofRWAur(Facchinietal.2016). Here,wehavediscoveredtwodimmingeventsinthe In any case, the variability during the dimming sug- WLTS V1334 Tau. The first fading occurred prior to gests an opacity change in the occulting feature. An- 2004,whereweobservea∼0.1magdimmingover∼300 other interpretation for the extended dimming is that days. The second event started in 2009 and has yet to a screen of hot dust can be created in the inner disk end. The photometric observations show a coherent andpushedacrossourline-of-sightbyhighdiskwinds dimming of ∼0.12 mag, an ingress dimming timescale (Petrovetal. 2015). However, we find no evidence for of ∼273days,andadurationof >2803days. Fromap- highorenhanceddiskwindsintheTRESspectra,mak- plyingasimplemodelwherethedimmingiscausedby ingthisscenariounlikely. Futuremillimetermappingof an occultation of V1334 Tau A by a large semi-opaque thediskaroundV1334Tauandradialvelocitymonitor- bodythathasaninclinedleadingedge,wedetermined ing for a short period companion will better constrain that the occulting body would have to be moving at thesescenarios. ∼14.2 kms−1,haveasemi-majoraxisof∼6.56AU,and be ∼24.3 AU in width. Assuming a circular orbit, this 6.3. IncreasedMagneticFieldFlux would suggest that the occulting featureis quite large, Another possible explanation of the dimming events asthewidthis∼59%ofitsentireorbit. Thiscouldbeex- seen in V1334 Tau is they are a result of significant in- plainedbyanasymmetricdiskordusttrapoccultingthe crease in the magnetic field flux at the surface of the star in a nearly edge-on orientated disk. Although we star. According to the MIST stellar evolution models, seenoIRexcess,thedouble-peakedHαlineseeninthe the interior structure of V1334 Tau is predicted to be TRESspectrasuggeststhepresenceoforbitingmaterial. fully convective. Currentmagneto-hydrodynamic stel- Itispossible thatthe occulting body maybe awarp or larstructuremodelsoffullyconvectivedwarfssuggest perturbationinthesurroundingdisk,butthiswouldre- thatthesestarsareabletosupportstable(>1000days), quireaclosetoedge-ongeometry. Itispossiblethatwe large-scale dipolar fields (> 1 kilogauss) despite rotat- areseeingafluctuationintheinnerdiskstructureorthat ing rapidly, resulting in the formation of large group- windshavepushedmaterialfromthediskplaneacross ingsofdarkpolarstarspotsblanketingasignificantfrac- our line of sight (Natta&Whitney 2000; Nelsonetal. tion of the stellar photosphere (Yadavetal. 2015). The 2000;Petrovetal.2015). However,wefindnoevidence formationoflong-livedpolarspotsmayresultinmulti- forhighdiskwindsinourTRESspectra. Wealsofinda yeardimmingeventssimilartotheonesseenhere,par- ∼0.32dayperiodicityintheKELTphotometrythatper- ticularlyifweareviewingthisstarnearlypole-on. The sists through the 2009 dimming. The specific cause of high amount of Hα emission, and the excess NUV and thisperiodicityisunclearbutitisconsistentwithorbit- FUV GALEX flux observed in the star’s SED, suggests ingmaterialonlyafewstellarradiifromthestar. that V1334 Tau is magnetically active. Future observa- Understanding the observed extended dimmings of tions should clarify if the dimming we observe is con- V1334Tauover the past∼15yearsrequiresthe contin- nected to magnetic activity. This includes stellar color ued photometric monitoring of the system. As a start, information in and out of dimming events, mapping V1334 Tau will be observed in the upcoming K2 cam- thestellarmagneticfieldsusingZeeman-DopplerImag- paign 13 (spring 2017). Simultaneous multi-band pho- ing,andmodelingtheupcomingK2highprecisionlight tometric and spectroscopic observations leading up to curves. and coinciding with the K2 observations may provide newinformationtodistinguish betweenpotentialdim- 6.4. SimilarSystems ming mechanisms. Future millimeter mapping of the There are only a dozen or so objects known to show V1334Tausystemshould detectandmeasurethe incli- thistypeofphenomena. ThedurationoftheV1334Tau nation of the possible circumstellar disk. Since V1334 dimming events strongly resemble the long term dim- Tau is a visual binary, millimeter mapping may also mingofthearchetypeAATau(Bouvieretal.2013). Al- show evidence of a stellar encounter, similar to what though similar in duration, the dimming event in AA hasbeenseeninRWAur(Cabritetal.2006). Upcoming Taushows a muchlargerdepthof ∼2mag. Other sys- largesurveysliketheTransitingExoplanetSurveySatel- temslike V409Tau, DM Ori, and RWAur have shown lite (TESS, Rickeretal. 2015), the Large Synoptic Sur- similar dimmings (again at much larger depths), and vey Telescope (LSST, LSSTScienceCollaborationetal. are interpreted as either a periodically obscuring fea- 2009), and the PLAnetary Transits and Oscillations of tureinthecircumstellardisk,highdiskwinds,orinner stars (PLATO, Catalaetal. 2010) mission should in- disk instabilities (Rodriguezetal. 2013, 2015, 2016d,c; crease our sample of these unique systems, allowing Petrovetal.2015;Facchinietal.2016).Itispossiblethat us to statistically address the variety in the depth and the dimming mechanisms for all these systems are re- timescalesofthesedimmingevents. lated. 7. SUMMARYANDCONCLUSIONS EarlyworkonKELT-NorthwassupportedbyNASA