Spectroscopy across the brown dwarf/planetary mass boundary - I. Near-infrared JHK spectra

3.9 MB·English

by  J. Patience

Astronomy&Astrophysicsmanuscriptno.astroph (cid:13)c ESO2012
January20,2012
Spectroscopy across the brown dwarf/planetary mass boundary - I.
Near-infrared JHK spectra(cid:63)
J.Patience1,R.R.King1,R.J.DeRosa1,A.Vigan1,S.Witte2,E.Rice3,Ch.Helling4,andP.Hauschildt2
1 AstrophysicsGroup,SchoolofPhysics,UniversityofExeter,Exeter,EX44QL,UK
e-mail:[email protected], [email protected], [email protected], [email protected]
2 HamburgerSternwarte,Gojenbergsweg112,21029Hamburg,Germany
e-mail:[email protected], [email protected]
3 Department of Astrophysics, American Museum of Natural History, 79th Street and Central Park West, New York, NY 10024,
USA;Currentaddress:DepartmentofEngineeringScience&Physics,CollegeofStatenIsland,2800VictoryBlvd,StatenIsland,
NY10314,USA
2 e-mail:[email protected]
1 4 SUPA,SchoolofPhysics&Astronomy,UniversityofSt.Andrews,NorthHaugh,St.Andrews,KY169SS,UK
0 e-mail:[email protected]
2
ReceivedSep.9,2011;acceptedJan.5,2012
n
a
ABSTRACT
J

8 WithauniformVLTSINFONIdatasetofninetargets,wehavedevelopedanempiricalgridofJ,H,Kspectraoftheatmospheres
1 ofobjectsestimatedtohaveverylowsubstellarmassesof∼5-20M andyoungagesrangingfrom∼1-50Myr.Mostofthetargets
Jup
arecompanions,objectswhichareespeciallyvaluableforcomparisonwithatmosphereandevolutionarymodels,astheypresentrare

] casesinwhichtheageisaccuratelyknownfromtheprimary.Basedontheyouthofthesample,allobjectsareexpectedtohavelow
P
surfacegravity,andthisstudyinvestigatesthecriticalearlyphasesoftheevolutionofsubstellarobjects.Thespectraarecompared
E withgridsoffivedifferenttheoreticalatmospheremodels.Thisanalysisrepresentsthefirstsystematicmodelcomparisonwithinfrared
. spectraofyoungbrowndwarfs.ThefitstothefullJHKspectraofeachobjectresultinarangeofbestfiteffectivetemperaturesof
h
±150-300Kwhetherornotthefullmodelgridorasubsetrestrictedtolowerlog(g)valuesisused.Thiseffectivetemperaturerangeis
p
significantlylargerthantheuncertaintytypicallyassignedwhenusingasinglemodelgrid.Fitstoasinglewavelengthbandcanvary
-
o byupto1000Kusingthedifferentmodelgrids.Sincetheoverallshapeofthesespectraisgovernedmorebythetemperaturethan
r surfacegravity,unconstrainedmodelfitsdidnotfindmatcheswithlowsurfacegravityoratrendinlog(g)withage.Thissuggeststhat
t empiricalcomparisonwithspectraofunambiguouslyyoungobjectstargets(suchasthosepresentedhere)maybethemostreliable
s
a methodtosearchforindicationsoflowsurfacegravityandyouth.Basedoncomparisonwithpreviousobservations,theSINFONI
[ spectrarepresentasecondepochforthetargets2M0141andDHTauB,andthecombineddatashownovariationsinthespectral

morphology over time. The analysis of two other targets, AB Pic B and CT Cha B, suggests that these objects may have lower
1
temperatures,andconsequentlylowermasses,thanpreviouslyestimated.
v
1 Keywords.planetarysystems,stars:browndwarfs,stars:atmospheres,binaries:close,techniques:highangularresolution
2
9
3 1. Introduction Theeffectsoflowsurfacegravityontheinfraredspectraof
.
1 brown dwarfs have been investigated with isolated members of
0 SincethediscoveryofthefirstsubstellarcompanionsGD165B star-formingregionsandyoungclusters.IntheJ-band,thealkali
2 (Becklin & Zuckerman 1988) and Gl 229B (Nakajima et al. metallinesofKIandNaIandtheFeHabsorptionwerefound
1 1995), there have been many surveys for additional low mass
tohavelowerequivalentwidthsinyoungbrowndwarfsrelative
: companions, but a very limited number have been discovered
v tofieldobjectsofthesamespectraltypeinaninitialsampleof
(e.g., McCarthy & Zuckerman 2004; Metchev & Hillenbrand
i threeyoungobjects(McGovernetal.2004)andalargersample
X 2006), particularly at the lowest masses (Zuckerman & Song
of 23 brown dwarfs in Upper Sco (Lodieu et al. 2008). Lower
r 2009). Substellar objects that are companions to nearby stars pressuresintheatmospheresoftheyoungobjectscausealower
a and brown dwarfs are especially valuable for comparison with
amountofrecombinationofKIIandNaIItoformKIandNa
atmosphereandevolutionarymodels,astheypresentrarecases
I which, along with pressure-broadening (Rice et al. 2010) and
in which the brown dwarf age is inferred by the accurate age
other processes, impacts the line strengths relative to the field
of the primary. Recent models (e.g., Fortney et al. 2008) have
objects, resulting in weaker lines. An index based on the Na I
indicated the importance initial conditions may have in deter-
doublet was defined to assess surface gravity in the youngest
mining the flux from young objects, making confirmed young
objects(Allersetal.2007).ThetriangularshapeoftheH-band
browndwarfsimportanttogaugetheeffectsofaccretionhistory was noted in the spectra of low luminosity σ Orionis members
andatmosphericprocessesatlowsurfacegravity.
(Lucas et al. 2001) and associated with a decrease in H col-
2
lision induced absorption due to low gravity in a young field
(cid:63) basedonobservationsobtainedattheParanalObservatory,Chile object (Kirkpatrick et al. 2006). Rice et al. (2011) demonstrate
for ESO programs 279.C-5010(A), 080.C-0590(A), 077.C-0264(A), how the relative opacities from CIA-H absorption and H O at
2 2
078.C-0800(B),&078.C-0800(A)
1
Patience,King,DeRosa,Vigan,Witte,Rice,HellingandHauschildt:Near-IRspectraacrossthebrowndwarf/planetboundary
different gravities shape the spectra of young objects. The CO
bandhead shape in the K-band has also been linked with lower
surfacegravityduetothesensitivityoftheabsorptiontomicro-
turbulence, and has been used to differentiate dwarf and giant
stars(Kleinmann&Hall1986).
Giventheintrinsicallylowluminositiesandtemperaturesof
substellar objects, infrared observations provide higher signal-
to-noise spectra than optical measurements, making this wave-
length range important for the characterization of the physical
properties of brown dwarfs. For example, J-band spectra taken
at two resolutions of a sample of M6-M9 objects including 15
browndwarfswithages∼1-10Myrwerecomparedwithagrid
ofmodelatmospherestoinferanumberofproperties–effective
temperatures, surface gravities, radial velocities and projected
rotationalvelocities(Riceetal.2010).
Thispaperpresentsthecombinedflux-calibratedJ,H,Kspec-
traforasetofyoung(2-50Myr),lowmass(5-25M )substel-
Jup
lar objects and compares the observations with a suite of syn-
theticspectradevelopedfromfivedifferenttheoreticalmodelat-
mospheresimulations.Therangeofphysicalpropertiesinferred
fromthedifferentmodelsisdiscussed,alongwiththediscrepan-
cies obtained from fits over portions of the J,H,K wavelength
range rather than the full combination. The spectral shape is
the focus of this paper, and the investigation of spectral lines Fig.1.Color-magnitudediagramofFieldM(grey),L(green),T(blue)
andlongerwavelengthscoveringtheL-bandarethesubjectofa objects,thesampletargets(filledred)andotheryoungcompanions(un-
forthcomingpaper(Kingetal.2011,inprep.). filledred),andtheHR8799planetswithmeasuredJandKmagnitudes
(black).ThefielddataweretakenfromtheDwarfArchive.organdre-
strictedtoobjectswithparallaxuncertainty<10%andcoloruncertainty
2. SampleandSummaryofPreviousObservations <0.5mag.ThevaluesusedforthetargetsaregiveninTables1and2.
For the young companions CHXR 73B (Luhman et al. 2006), 1RXS
The sample consists of 9 objects estimated to have very low
J1609B (Lafrenie`re et al. 2008), GSC 06214 B (Ireland et al. 2011)
massesof∼5-20MJupandyoungagesrangingfrom∼1-50Myr, and HD 203030B (Metchev & Hillenbrand 2006) that are plotted on
placingthemattheboundaryofthebrowndwarfandplanetary thefigure,butnotincludedinthisstudy,thedataweretakenfromthe
massregimes.Thesamplecoversthreeagebins,withthree∼1- discoverypapers.
3 Myr old members of star-forming regions, three members of
the∼8MyrTWHydraAssociation,andthreetargetswithages
Table2.Photometry
of∼30-50Myr.Eightofthetargetsaremembersofbinarysys-
tems discovered with adaptive optics imaging surveys, and one
Source J H K
object is an isolated young brown dwarf. The youngest objects
CTCha A 9.72±0.02 8.94±0.05 8.66±0.02
may still be surrounded by material from their formation envi- B 16.6±0.3 ... 14.9±0.3
ronment and have the possibility of accretion and disks, while DHTau A 9.77±0.02 8.82±0.03 8.18±0.03
the older objects likely present bare photospheres. All are ex- B 15.71±0.05 14.96±0.04 14.19±0.02
pectedtohavelowsurfacegravity.Thedistinctposition(dered- GQLup A 8.69±0.04 7.70±0.03 7.10±0.02
dened,whennecessary)onthecolor-magnitudediagramofthis B 14.90±0.11 ... 13.34±0.13
young, low surface gravity sample compared to field objects is
showninFigure1;thetrendisaspredictedintheoreticalevolu- 2M1207 A 13.00±0.03 12.39±0.03 11.95±0.03
B 20.0±0.2 18.09±0.21 16.93±0.11
tionarymodels(e.g.,Saumon&Marley2008).Thetargetsshow
TWA5 Aa 8.40±0.07 7.69±0.04 7.39±0.04
reddercolorsthanfieldLandTdwarfs,analogoustotheimaged
Ab 8.5±0.2 7.79±0.05 7.62±0.08
planetsorbitingHR8799.Forallmembersofthesample,there
B 12.6±0.2 12.14±0.06 11.4±0.2
are archive infrared J,H,K spectra from the VLT integral field
spectrographSINFONI,formingauniformdatasetwithwhich ABPic A 7.58±0.03 7.09±0.03 6.98±0.03
it is possible to construct an empirical grid of the atmospheres B 16.18±0.10 14.69±0.10 14.14±0.08
of substellar objects covering the critical early phases of their GSC08047 A 9.06±0.03 8.53±0.06 8.41±0.03
evolution. B 15.9±0.1 15.45±0.02 14.8±0.1
2M0141 14.83±0.04 13.88±0.03 13.10±0.03
2.1. Star-formingregionsystems
High angular resolution imaging surveys targeting members of
theneareststar-formingregionshaveidentifiedanumberofvery AU at a distance of 150 ± 20 pc (Franco 2002), and has been
young(∼1-2Myr)wideorbit(>100AU)lowmassbrowndwarf the subject of considerable prior investigation into the inferred
companions.Thepre-MainSequencesampleconsideredforthis companionmassandevidenceforongoingaccretion.Basedon
paper are the three companions with archive SINFONI J,H,K a comparison of theoretical models with photometry and K-
spectra: GQ Lup B (Neuha¨user et al. 2005), DH Tau B (Itoh bandspectroscopy,themasswasinitiallyestimatedas1-42M
Jup
et al. 2005), and CT Cha B (Schmidt et al. 2008). The clos- (Neuha¨user et al. 2005), while subsequent analysis of a larger
estcompanionisGQLupB,withaseparationof0(cid:48).(cid:48)7,or∼105 wavelength range (e.g., Marois et al. 2007) revised the mass
2
Patience,King,DeRosa,Vigan,Witte,Rice,HellingandHauschildt:Near-IRspectraacrossthebrowndwarf/planetboundary
Table1.Sample
Source Primary Comp. Age M M D Sep. Membership A Refs.
prim sec J
(Myr) (M ) (M ) (pc) (AU)
(cid:12) Jup
GQLupB K7 L1.5 1 0.7 24±12 150±20 100 LupusSFR 0.1±0.2 1-5
DHTauB M0.5 L2 1 0.33 11±10 140±15 330 TaurusSFR 0.3±0.3 6,2
3
CTChaB K7 ≥M8 2 0.7 17±6 165±30 440 ChameleonSFR 0.4±0.2 7-8
2M1207A&B M8 L5 8 25±5M 6-10 52.4±1.1 46 TWHydraAssoc. – 9-12
Jup
TWA5B M1.5 M8/8.5 8 0.40 ∼20 44±4 98 TWHydraAssoc. – 13-15
ABPicB K2 L0-L1 30 0.84 13-14 47.3±1.8 248 Tuc-HorAssoc. – 16-17
GSC08047B K3 M9.5 30 0.80 25±10 ∼85 279 Tuc-HorAssoc. – 18-19
2M0141 L0 – 1-50 15±10M – ∼35 – Tuc-Hor/βPic? – 20
Jup
References: (1)Neuha¨useretal.(2005)(2)Luhmanetal.(2006)(3)Maroisetal.(2007)(4)Franco(2002)(5)Batalhaetal.(2001)(6)Itohetal.
(2005)(7)Schmidtetal.(2008)(8)Luhman(2004)(9)Gizis(2002)(10)Chauvinetal.(2004)(11)Chauvinetal.(2005a)(12)Mohantyetal.
(2007)(13)Lowranceetal.(1999)(14)Webbetal.(1999)(15)Konopackyetal.(2007)(16)Chauvinetal.(2005c)(17)Bonnefoyetal.(2010)
(18)Chauvinetal.(2005b)(19)Neuha¨user&Guenther(2004)(20)Kirkpatricketal.(2006)
range to 10-20M . In Table 1, the mass is listed as 24 ± 12 2000), the H- and K-band spectra have been published sepa-
Jup
M (Luhmanetal.2006),sincethisestimateappliedthesame rately (Neuha¨user et al. 2009). The second TW Hydra mem-
Jup
approachtobothGQLupBandanothertargetDHTauB.The ber, 2M1207, is composed of two substellar objects (Chauvin
SINFONIspectrareportedinSeifahrtetal.(2007)suggestedev- et al. 2004, 2005a) – the primary is an M8 brown dwarf (Gizis
idence for active accretion from the Pa β emission line, while 2002) similar to TWA 5B and the secondary is a L-type plan-
independentJ,H,KspectrafromOSIRIS(McElwainetal.2007) etary mass object (Chauvin et al. 2004; Mohanty et al. 2007).
and NIFS (Lavigne et al. 2009) do not confirm the line, which ThecompleteJ,H,Kspectrumof2M1207Bhasbeenpublished
may have resulted from contamination from the primary.The (Patienceetal.2010),butnotthecorrespondingspectrumofthe
SINFONIspectrumofGQLupisincludedinthispaperforcom- primary.WhencomparedwithDUSTYmodels,thecompanion
pleteness. spectrumpresentsaspectralshapeconsistentwithtemperatures
The companion DH Tau B (Itoh et al. 2005) is in a wider higherthanexpectedfromtheluminosity(Mohantyetal.2007;
orbit,at2(cid:48).(cid:48)3(330AU)fromtheprimary,makingcontamination Patience et al. 2010). The discrepancy between spectral shape
fromthehoststarunlikely.Theestimatedmassforthecompan- andfluxlevelwasinitiallyexplainedbyobscurationfromagrey
ion, 11+10M , listed in Table 1 is based on the same model disk (Mohanty et al. 2007), and then updated to include mod-
−3 Jup
comparison applied to GQ Lup B (Luhman et al. 2006) and elsincorporatinganimprovedtreatmentofclouds(Skemeretal.
withinthe5-40M rangereportedinItohetal.(2008).Thepre- 2011)andbothcloudsandnon-equilibriumchemistry(Barman
Jup
viousCISCOJ,H,Kspectrum(R∼440)showedabsorptionfrom et al. 2011) to simultaneously fit the spectrum shape and flux
H O bands indicating a low effective temperature, and no evi- level. The 2M1207 B spectrum represents an example of how
2
dence of emission lines from accretion was present (Itoh et al. measurements of young, low mass object atmospheres can ad-
2005).TheSINFONIdatapresentedinthispaperweretakenat vancetheunderstandingoftheatmosphericphysics.
higher resolution (R∼1500) than the CISCO spectrum and at a
differentepoch,enablingasearchforvariability.
The widest young companion CT Cha B is 2(cid:48).(cid:48)7 (440 AU) 2.3. Tucana-Horolgiumsystems
from the primary (Schmidt et al. 2008). Based on fitting the
An additional two substellar companions – AB Pic B and
SINFONIJ,H,KspectrawithDrift-PHOENIXmodelstodeter-
GSC 080647 B – are members of the older ∼30 Myr Tucana-
minethetemperatureandphotometrytoestimatetheluminosity, Horologium Association. AB Pic B is a wide orbit 5(cid:48).(cid:48)5 (∼260
thecompanionmasswasestimatedtobe17±6M (Schmidt
Jup AU)commonpropermotioncompanion(Chauvinetal.2005c)
etal.2008)fromcomparisonwithevolutionarymodels(Baraffe
toaK2primary(Perrymanetal.1997).Basedonacomparison
etal.2003;Chabrieretal.2000;Burrowsetal.1997).Thespec-
of the system photometry with evolutionary models (Burrows
trum shows Pa β emission, an indication of ongoing accretion, etal.1997;Baraffeetal.2002)at30Myr,themassofthecom-
though another accretion diagnostic Br γ is not seen (Schmidt
panion is estimated to be 13-14M (Chauvin et al. 2005c),
et al. 2008). In this paper, we reconsider the SINFONI spectra Jup
placing AB Pic B at the demarcation of planetary masses and
in comparison with a larger set of models and with a different
brown dwarf masses. The J,H,K SINFONI spectra have been
treatmentofextinction.
used to estimate a spectral type for the companion of L0-L1
and to compare each band of the spectrum with theoretical at-
mospheremodels(Bonnefoyetal.2010).Inthispaper,amodel
2.2. TWHydraAssociationsystems
comparisonofthecombinedflux-calibratedJ-Kspectrumispre-
Two members of the TW Hydra Association – TWA 5 and sented, and the combined spectrum shows differences with the
2M1207–includethreesubstellarobjects,withbothlate-Mand models not apparent in the single band comparisons. Another
mid-L spectral types. TWA 5B was one of the earliest brown Tucana-HorologiumpairwithverysimilarspectraltypestoAB
dwarfcompanionsidentified(Lowranceetal.1999;Webbetal. Pic is GSC 08047, which is composed of a K3 primary with a
1999) and is located 2(cid:48).(cid:48)0 (∼100 AU) from a close pair of early confirmed substellar companion ∼3(cid:48).(cid:48)3 (279 AU) from the host
M-stars with a ∼5 yr orbit (Konopacky et al. 2007). Although star (Chauvin et al. 2005b). Previous spectra in the H- and
thephotometryofTWA5Bhasbeenexplored(Neuha¨useretal. K-bands taken with VLT/ISAAC and VLT/NACO were com-
3
Patience,King,DeRosa,Vigan,Witte,Rice,HellingandHauschildt:Near-IRspectraacrossthebrowndwarf/planetboundary
Table3.SINFONIObservations
Source Grating Scale NDIT DIT #exp. Dates ESOProjectID PI
CTChaB J 100mas 2 138.5s 6 18May07 279.C-5010(A) Schmidt
CTChaB H+K 100mas 2 138.5s 6 16May07 279.C-5010(A) Schmidt
DHTauB J 25mas 1 300s 32 18Dec07 080.C-0590(A) Rojo
DHTauB H+K 25mas 1 300s 8 22Oct07 080.C-0590(A) Rojo
GQLupB J 25mas 1 300s 27 18Sep06 077.C-0264(A) Neuha¨user
GQLupB H 25mas 1 300s 11 24Ap06 077.C-0264(A) Neuha¨user
GQLupB K 25mas 1 300s 8 17Sep05 077.C-0264(A) Neuha¨user
TWA5B J 25mas 1 120s 6 12Dec07 080.C-0590(A) Rojo
TWA5B H+K 25mas 2 60s 7 13Dec07 080.C-0590(A) Rojo
2M1207A+B J 100mas 1 600s 22 25Feb07 078.C-0800(B) Thatte
2M1207A+B H+K 100mas 1 300s 24 28Jan07,07Feb07 078.C-0800(B) Thatte
ABPicB J 25mas 1 300s 40 5-7Dec07 080.C-0590(A) Rojo
ABPicB H+K 25mas 1 300s 8 11Nov07 080.C-0590(A) Rojo
GSC08047B J 25mas 1 300s 24 6&18Jan08 080.C-0590(A) Rojo
GSC08047B H+K 25mas 1 300s 27 5&9Jan08 080.C-0590(A) Rojo
2M0141 J 250mas 1 300 4 19Oct06 078.C-0800(A) Thatte
2M0141 H+K 250mas 1 60s 4 19Oct06 078.C-0800(A) Thatte
pared with late-M objects to estimate a spectral type of M8 ± companions are separated by less than a few arcseconds from
2 (Neuha¨user & Guenther 2004) and M9.5 ± 1 (Chauvin et al. theprimaries,makingtheprimariesidealsourcesfortheAOcor-
2005b), slightly earlier than the AP Pic companion. The new rection.Theobservingsequenceforalltargetsincludedasetof
SINFONI spectra include the J-band and the IFU observations exposures on the target with offset positions on the array. For
arenotsubjecttowavelength-dependenteffectsofdifferencesin sometargets,separate blankskyexposureswerealso recorded.
theslitcenteringfortargetandstandardthataffectedsomeofthe Standardcalibrationobservationsofanearly-typeorsolar-type
previousspectra. starfollowedeachtargetsequencetocorrectfortelluricfeatures
and the instrument response. Additional calibrations were ob-
tained to measure the wavelength scale, the detector dark cur-
2.4. YoungIsolatedFieldObject rent,distortion,andquantumefficiencyvariations.Thedatafor
Aspartofthelargescalesearchandcharacterizationprogramfor this project were drawn from the ESO archive, and a summary
L-dwarfswith2MASS,anisolatedbrowndwarfwithindications of the instrument configuration, observation details and project
ofyouthwasidentifiedwithspectroscopyspanning0.6-2.5µm IDcodesforeachtargetisgiveninTable3.Insomecases,only
(Kirkpatrick et al. 2006). In the infrared, there is a J-K SpeX thesubsetofhigherqualitydatawereused;Table3listsonlythe
spectrumwithsimilarresolution(R∼1200)totheSINFONIdata, observationsusedinthisanalysis.
andaNIRSPECJ-bandspectrumwithhigher(R∼2500)resolu- There are four gratings available in SINFONI – J, H, K,
tion,making2M0141anidealcasetotestforvariabilitysuchas andH+K–whichproducespectralresolutionsofapproximately
hasbeenreportedforTWA30(Looperetal.2010).For2M0141, 2000,3000,4000,and1500,dependingonthescale.Foralltar-
the spectral features of weak alkali lines such as K I in the J- getsexceptGQLupB,acombinationofJandH+Kobservations
band and the triangular shape of the H-band caused by a de- were taken. For GQ Lup B, separate J, H, and K spectra were
creaseinH collisioninducedabsorptionprovidedevidenceofa recorded.Amongthethreeavailablespaxel(spatialpixel)scales
2
lowsurfacegravityobject(Kirkpatricketal.2006).Theeffective of25mas,100mas,and250mas,the100masscalewasemployed
temperatureandsurfacegravitywerederivedfromfitstoatmo- for the CT Cha B and 2M1207A+B observations. The 100mas
sphericmodelsgeneratedwiththePHOENIXcode(Kirkpatrick scaleprovidesafield-of-viewof3(cid:48).(cid:48)0×3(cid:48).(cid:48)0,andboththebrown
etal.2006),andTeffandlog(g)werecomparedwithevolutionary dwarfprimaryandplanetarymasssecondaryfor2M1207were
models (Baraffe et al. 2002) to infer an age range of 1-50 Myr included within each observation. For the remaining target, the
andobjectmassof6-25Myr(Kirkpatricketal.2006).Although spaxelscaleof25maswasusedandthefield-of-viewatthisscale
2M0141 currently has a less certain age, future measurements is0(cid:48).(cid:48)8×0(cid:48).(cid:48)8whichexcludedtheprimaryhoststarfromtheob-
oftheparallaxandspacemotionmaydeterminemembershipin servations.Foronetarget,GQLupB,thediffractionspikefrom
theTucana-HorologiumAssociationortheβPicmovinggroup. theprimaryintersectedthepositionofthetargetcompanionina
In this paper, the SINFONI spectra are compared with a larger subsetoftheobservations.
setofatmospheremodelsandwiththeearlierSpeXspectrumto
searchforvariability.
4. DataAnalysis
3. SINFONIObservations 4.1. SINFONIdatareduction
The observations were taken with the SINFONI (Spectrograph The initial steps of the data reduction for all targets were per-
for INtegral Field Observations in the Near Infrared) instru- formedwiththeGasgano(Gebbincketal.2007)implementation
ment (Bonnet et al. 2004; Eisenhauer et al. 2003; Thatte et al. oftheSINFONIdatareductionpipeline(Modiglianietal.2009;
1998), an AO-equipped integral field spectrograph mounted at Dumasetal.2007).Theprocessingbeganwiththerawdataand
the Cassegrain focus of the VLT (UT4). The target low mass calibrationfilesintheESOarchiveandisindependentofprevi-
4
Patience,King,DeRosa,Vigan,Witte,Rice,HellingandHauschildt:Near-IRspectraacrossthebrowndwarf/planetboundary
Fig.2.Theevolutioninsurfacegravityoverthe1-100Myragerange Fig.3. The extinction A of the secondary as a function of the A
J J
forobjectswithmassesof∼1-50M .Thevaluesaretakenfromthe of the primary for spatially resolved binary systems in Lupus (red),
Jup
DUSTY00models(Chabrieretal.2000)andshowthesystematicin- Chameleon(lightblue),andTaurus(darkblue)basedonobservations
creaseinlog(g)overtimeastheobjectscontract.Thesetracksareused of the individual components of each system (Brandner & Zinnecker
to determine the restricted range of surface gravity to allow in model 1997;White&Ghez2001;Pratoetal.2003).Thesebinarieshavecom-
fitstoouryoungtargets. parable separations to the target systems and are located in the same
star-formingregions.Thedatashowatrendofcomparableextinction
foreachcomponent.
ously reported results summarised in Section 2. This approach
was taken to ensure that all data used the same software ver-
sion and were reduced in a consistent way. Corrections for the moved by multiplication of a black body with the same effec-
dark current, linearity response, and bad pixels were measured tivetemperatureasthestandard.Tofluxcalibratethespectra,it
and applied to all data files. The instrument optical distortion was necessary to rely on photometric measurements presented
was measured with a set of fibers illuminating positions across in Table 2 which were not contemporaneous with the spectra.
thefieldandthenon-linearwavelengthdispersionwasmeasured TheobservedspectraandthespectrumofVega(Mountainetal.
with observations of an arc lamp; both the spatial and spectral 1985; Hayes 1985) were convolved with the 2MASS response
calibrationswereappliedtothedatawithGasganoroutines.The curves(Carpenter2001)tocalculatethescalingfactorrequired
calibrated two-dimensional data for each observation was con- to match the reported magnitude in a given filter. For the H+K
vertedintoathree-dimensionaldatacubeconsistingofanimage spectra, the H-band photometry was used for flux calibration,
at each wavelength slice. Across the entire wavelength range, becausetheSINFONIdatacoversthefull2MASSH-bandfilter
a stack of ∼700-2000 wavelength slices images were recorded, bandpass.Foronetarget,GQLupB,thereisnoreportedH-band
howeverthefaintesttarget2M1207BrequiredbinningintheJ- magnitudeandtheHandKspectrawereobservedseparately.In
bandtoproduceaspectrumwithahighersignal-to-noiseratio. thiscase,theH-bandmagnitudewasestimatedbyassumingthe
After corrections for the detector and optics, the sky back- same H-K color as DH Tau B, since GQ Lup B and DH Tau
ground emission was subtracted from the offset position cover- B both have a very similar J-K color and similar young ages.
ingeitherblankskyorincludingthetargetatadifferentposition Another target, CT Cha B, also had no H-band measurement,
onthearrayintheABBApatternofobservations.Theindividual and, in this case, the K-band magnitude was used to flux cali-
sky-subtractedcubeswerealignedbasedoncentroidingthetar- bratetheH+Kspectrum.
getinthetopwavelengthsliceandthencombined.Toensurethe
finalspectrumtracedastraightpaththroughthedatacube,each
4.2. Comparisonwiththeoreticalmodelgrids
wavelengthsliceofthefinaldatacubewasshiftedandaligned,
andthenthefinalspectrumwasextracted.Fortheonetargetwith The target spectra were compared with five sets of theoreti-
bothcomponentsonthearray,2M1207,itwasnecessarytosub- cal models, spanning a range of effective temperature and sur-
tractthehalooftheprimarybeforeextractingthespectrumofthe facegravity,asdescribedinSection5.Themodelspectrawere
companion.Furtherdetailsofthesubtractionhavebeenreported smoothedtothesameresolutionasthedataandinterpolatedto
previously(Patienceetal.2010). thesamedispersionbeforeperformingafittingprocedureusing
The standard stars observed after the targets were used to the least-squares statistic defined in Mohanty et al. (2007) and
correct for telluric absorption and the wavelength response of allowing the model flux to be scaled to find the best fit. This
the detector by dividing each target spectrum by its associated fluxscalingisrelatedtotheobjectradius,enablinganestimate
standard after interpolating the standard spectrum over intrin- of the radius in addition to the effective temperature and sur-
sic features. The shape of the standard star continuum was re- facegravity.SinceeachwavelengthstepintheSINFONIspec-
5
Patience,King,DeRosa,Vigan,Witte,Rice,HellingandHauschildt:Near-IRspectraacrossthebrowndwarf/planetboundary
10 5.5 6
9 Teff = 1400K 5 Teff = 1700K Teff = 2000K
8 4.5 5 DUSTY
7 4
4
6 3.5 BT(cid:239)Settl
3
f(cid:104) 5 f(cid:104) 2.5 f(cid:104) 3 DRIFT
4
2
2
3 1.5 GAIA(cid:239)Dusty
2 1 1
1 0.5 Marley et al.
0 0 0
1 1.2 1.4 1.6 1.8 2 2.2 2.4 1 1.2 1.4 1.6 1.8 2 2.2 2.4 1 1.2 1.4 1.6 1.8 2 2.2 2.4
Wavelength (µm) Wavelength (µm) Wavelength (µm)
6 5 5
Teff = 2300K 4.5 Teff = 2600K 4.5 Teff = 2900K
5
4 4
3.5 3.5
4
3 3
f(cid:104) 3 f(cid:104) 2.5 f(cid:104) 2.5
2 2
2
1.5 1.5
1 1
1
0.5 0.5
0 0 0
1 1.2 1.4 1.6 1.8 2 2.2 2.4 1 1.2 1.4 1.6 1.8 2 2.2 2.4 1 1.2 1.4 1.6 1.8 2 2.2 2.4
Wavelength (µm) Wavelength (µm) Wavelength (µm)
Fig.4.Forafixedvalueoflog(g)=4.5,thesolarmetallicitymodelspectraattemperaturesrangingfrom1400K(upperleft)to2900K(lowerright)
instepsof300K,correspondingtothetargeteffectivetemperatures.Fromtoptobottomineachpanel,themodelsare:DUSTY(grey),BT-Settl
(green),Drift-PHOENIX(red),Gaia-Dusty(purple),andMarleyetal.(blue).TheMarleyetal.modelsdonotincludetemperaturesabove2400K,
andtheplotsshowmodelswithf =1,thedustiestsetofmodelswithinthismodelgrid.
sed
Table4.CloudModelsa
Physics DUSTY BT-Settl Drift-PHOENIX Gaia-Dusty Marleyetal.
Opacitysource 30dustspecies 43dustspecies 7dustspecies updatedDUSTY 5dustspecies
Gas-dustphaseb S =1 S =1.001 S = f(z,s) S =1 S =1
Mixing none timescalefromRHD dampingofturnovertimescale none diffusivec
Settling/Rainout none timescalecomparison equationofmotion none parameterisedinf
sed
Dustsizes f(a)∝a−3.5,(ISM) timescalecomparison f(a,z)d f(a)∝a−3.5,(ISM) log-normal f(z,s)e
0.00625-0.24µm 0.00625-0.24µm 10−6−100µm 0.00625-0.24µm 10−5−1µm
Chemistry equilibrium equilibrium equilibriumf equilibriumf equilibriumexcept
CO/CH ,N /NH
4 2 3
Notes:aPHOENIX(Hauschildtetal.1999)DUSTY(Allardetal.2001),BT-Settl(Allardetal.2003,2010),Drift-PHOENIX(Hellingetal.
2008;Witteetal.2009),Gaia-Dusty(Riceetal.2010),Marleyetal.(Ackerman&Marley2001;Marleyetal.2003;Saumon&Marley2008;
Freedmanetal.2008)bS quantifiesthesupersaturationratio.S =1forthermalequilibriumbetweenthegasanddust.cParameterisedink .c
zz
f(a,z)isthegrainsizedistributionatheightzandprovidesthenumberofgrainsofsizea.e f(z,s)isthegrainsizedistributionforagrainspecies
satheightz. f UpdatedchemistrytreatmentcomparedtoDUSTY.
traisnearlyconstantovertheJHKrange,wedonotemploythe shown as a function of primary A in Figure 3 for binary stars
J
weighting techniqueof Cushinget al.(2008) whichwas devel- inthesameregionsasthetargets(Brandner&Zinnecker1997;
opedforspectracoveringamuchlargerrangeofwavelengths. White&Ghez2001;Pratoetal.2003),andacleartrendofsim-
ilarextinctionvaluesisseen.Basedonthebinarystars,thetar-
Two sets of fits were performed to the model grids – a fit
gets in this sample are assumed to have the same extinction as
including the full range of log(g) values and a fit restricted to
theprimary,whichisreportedinTable1.
log(g)≥3.0andwithanupperlimitbasedonevolutionarymodels
(Chabrieretal.2000)showninFigure2.Thetargetshaveknown
young ages, so the predicted evolution in surface gravity was
used to set log(g) limits: 4.0 for the companions to pre-Main 5. TheoreticalModels
Sequencestars,4.5fortheTWHydramembers,and5.0forthe
The data were compared with five grids of theoretical atmo-
30-50Myrtargets.
spheremodels.Fourofthemodelsrepresentdifferentextensions
A final factor affecting the slope of the observed spectrum ofthePHOENIX(Hauschildtetal.1999)codeandincorporatea
is the amount of extinction to each target. This is particularly rangeofdustpropertiesandphysicaleffects.Inchronologicalor-
importantforthethreeobjectsinstar-formingregions.Noneof derofdevelopment,thePHOENIX-basedmodelsconsideredin
thesecondarieshavemeasuredvaluesofextinction,however,the thispaperinclude:DUSTY(Allardetal.2001),BT-Settl(Allard
primaries have reported A values. Rather than treat extinction et al. 2003, 2010), Drift-PHOENIX (Helling et al. 2008; Witte
V
asafreeparameter,weusetheresultsofbinarystarcomponent et al. 2009), and Gaia-Dusty (Rice et al. 2010), with the latter
extinctioncomparisonstogaugethelikelihoodofverydifferent threerepresentingongoingmodellingefforts.Theearliergenera-
A values for the primary and secondary. The secondary A is tionDUSTYmodelisincludedforcomparison,althoughtheBT-
J J
6
Patience,King,DeRosa,Vigan,Witte,Rice,HellingandHauschildt:Near-IRspectraacrossthebrowndwarf/planetboundary
Table5.ModelGridParameters 4
3.5
Model Teff log(g) ∆λ
range step range step 3
DUSTY 500-4000 100 3.5-6.0 0.5 2Å 2.5
BT-Settla 400-3000 100 3.5-5.5 0.5 0.2Å f(cid:104) 2
Drift-PHOENIX 1000-3000 100 3.0-5.5 0.5 2Å
1.5
Gaia-Dusty 1400-3500 50 3.0-6.5 0.1 0.5Å
1
Marleyetal.b 900-2400 100 4.0-5.0 0.25 R∼50,000
Note:aforTeff=2000-3000K,thelog(g)rangeis0.0-5.5bthe 0.5
wavelengthstep∆λisafunctionofλ 0
1 1.2 1.4 1.6 1.8 2 2.2 2.4
Wavelength (!m)
3.5
SettlandGaia-DustymodelshavesupersededtheDUSTYmod-
3
els with improved treatments of abundances and line lists. The
independent model grid developed by Marley et al. (Ackerman 2.5
& Marley 2001; Marley et al. 2003; Saumon & Marley 2008; (cid:104) 2
f
Freedmanetal.2008)isalsocomparedwiththetargetspectra.
1.5
A summary of the physics incorporated into the five mod-
els is given in Table 4; details of the models are given in the 1
referenceslistedaboveandthecomparisonofmodelswithtwo 0.5
benchmark test cases is reported in (Helling et al. 2008). The
0
1 1.2 1.4 1.6 1.8 2 2.2 2.4
treatmentofsomephysicalprocessessuchasconvectionisiden- Wavelength (!m)
ticalforallmodels,whiletheapproachtootherfactorssuchas 4.5
condensate settling and chemistry varies between the models. 4
Since dust plays a large role in determining the temperature- 3.5
pressurestructureoftheatmospheresinthetemperatureregime
3
appropriateforthetargets,thenumberofdustspeciesanddust
(cid:104) 2.5
size distributions involved in the calculations are listed, along f
2
withnotesonthemixingandsettlingincludedinthemodels.
1.5
Anexampleofthesyntheticspectraproducedbyeachmodel
over the same range of Teff for a fixed value of log(g) = 4.5 is 1
giveninFigure4.Thisvalueoflog(g)wasselectedbecauseitis 0.5
within the range predicted by evolutionary models for the ages 0 1 1.2 1.4 1.6 1.8 2 2.2 2.4
Wavelength (!m)
of the targets. For the Marley et al. spectra, the value of f is
sed
1whichrepresentsthedustiestatmospheres.Atthecoolesttem- Fig.5.SINFONIJ,H,Kspectraofthesample,separatedintothreeage
perature shown, 1400 K, there is a significant difference in the groups:(toppanel,toptobottom)the1-3MyrobjectsDHTauB,GQ
syntheticshapeoftheH-bandspectrumandtheoverallslopefor LupB,andCTChaB;(middlepanel.toptobottom)the8Myrobjects
the different models. These effects are reduced by 1700K, but 2M1207A,TWA5Band2M1207B;(bottompanel,toptobottom)the
30-50MyrobjectsGSC08047B,2M0141,andABPicB.Foreachage
stillimpactthepeakoftheJ-band.Byatemperatureof2000K,
group,thespectraareorderedfromhottest(top)tocoolest(bottom),and
theslopeacrosstheJ-bandrangesfromflattoincreasing.Athot- anoffsethasbeenaddedtothefluxforeaseofcomparison.Thespectra
tertemperatures,thelinestrengthsdifferentiatethemodelsmore
of the youngest subset have been dereddened based on the extinction
than the overall shape, and the atomic and metal hydride lines measuredfortheprimaries.
willbeinvestigatedforthissampleinafuturepaper(Kingetal.
2011,inprep.).
These model differences are primarily caused by the differ- induced when the convolution with a Gaussian instrument dis-
enttreatmentsofcondensatesettlingandthedifferencesindust persionprofileisundersampled,resultinginlowerfidelitymodel
opacity.Typically,ahigheroveralldustopacityresultsinarel- spectraforDUSTYandDrift-PHOENIX.Thenumericalorigin
ativelyloweredJ-bandfluxandweakensthewaterbandswhile of the difference was confirmed by calculating one grid (Gaia-
smoothing their edges. The H band shape is dominated by the Dusty) with two initial resolutions and comparing the results.
two adjacent water bands and the collision induced absorption Due to the fact that the model fits in this paper are based only
(CIA)whichbothgrowstrongerwithdecreasingdustopacities on the spectral shape, the additional width in the DUSTY and
asfluxemergesfromthedeeper,hotteratmosphereresultingin Drift-PHOENIXmodelsdoesnotimpactthisanalysis,however
abluerspectrum. thefuturecomparisonofspectrallinesandindicesrequiresmod-
Another difference between the models is apparent upon elswithsufficientresolutiontofitlinestrengths.
close comparison of the strong absorption lines and bands be- Table5showstherangeofTeff andlog(g)coveredbytheso-
tween models calculated at different resolutions. In regions of lar metallicity models. The Marley et al. grids also incorporate
strong,variedabsorption,modelscalculatedatlowerresolution aparameterf describingthesedimentationefficiency,andthe
sed
(2 Å) look ‘noisier’, i.e. a larger spread in the flux values over valuesconsideredforf inthisgridare1and2,withthehigher
sed
smallwavelengthranges(seeFig.4).Forexample,theDUSTY value associated with more efficient sedimentation from more
models show the largest ‘noise’ of the spectrum over the full rapid particle growth and larger average particle sizes. In addi-
temperature range, while the Marley et al., BT-Settl, and Gaia- tion to the solar metallicity models summarized in Table 5, the
Dusty models are typically much narrower in the range of flux Drift-PHOENIX grid also covered non-solar metallicities with
atagivenwavelength.Thesedifferencesareanumericalartifact values of -0.6, -0.3, 0.0, and +0.3. The data are first compared
7
Patience,King,DeRosa,Vigan,Witte,Rice,HellingandHauschildt:Near-IRspectraacrossthebrowndwarf/planetboundary
Table6.BestFittingAtmopshereModelParameters,FullGridComparison
WavelengthRange DUSTY BT-Settl Drift-PHOENIX Gaia-Dusty Marleyetal.
Teff (K) logg Teff (K) logg Teff (K) logg Teff (K) logg Teff (K) logg fsed
1-3MyrObjects
DHTauBa
JHK(1.10–2.46µm) 2600 4.0 2400 2.5 2000 4.0 2350 3.7 ≥2400b 4.0 2
J(1.10–1.34µm) 2400 4.0 2200 3.5 1600 3.5 2350 4.1 ≥2400b 4.0 2
HK(1.46–2.46µm) 2600 3.5 2500 3.5 1900 5.5 2550 3.8 ≥2400b 4.0 2
GQLupBa
JHK(1.10–2.46µm) 1900 6.0 1800 5.5 1700 3.5 2050 5.5 1600 4.5 1
J(1.10–1.34µm) 2500 4.0 2400 3.5 2600 4.0 2500 4.8 ≥2400b 4.0 2
HK(1.46–2.46µm) 1800 6.0 1700 4.5 1700 5.0 1800 5.5 1500 4.5 1
CTChaBa
JHK(1.10–2.46µm) 1800 4.5 1600 4.0 1700 3.5 1950 4.1 1500 4.5 2
J(1.10–1.34µm) 1900 3.5 1400 5.0 1700 3.0 2000 3.9 1100 5.0 2
HK(1.46–2.46µm) 1900 6.0 2200 1.0 1800 5.0 1950 5.5 1700 4.5 1
8MyrObjects
2MASS1207A
JHK(1.10–2.46µm) 3100 5.5 2800 5.0 ≥3000b 5.0 2900 5.3 ≥2400b 5.0 1
J(1.10–1.34µm) 2900 5.0 2700 4.5 ≥3000b 5.0 2850 5.1 ≥2400b 4.0 2
HK(1.46–2.46µm) 3100 5.5 2800 5.5 ≥3000b 5.5 2950 5.5 ≥2400b 5.0 1
TWA5B
JHK(1.10–2.46µm) 2500 3.5 2400 2.5 2200 3.0 2450 3.3 ≥2400b 4.0 2
J(1.10–1.34µm) 2700 4.5 2600 4.0 2800 4.5 2700 5.0 ≥2400b 4.0 2
HK(1.46–2.46µm) 1800 6.0 2200 1.0 1800 5.5 1900 5.5 1600 4.5 1
2MASS1207B
JHK(1.10–2.46µm) 1600 3.5 1500 3.5 1500 5.0 1650 4.3 1100 5.0 1
J(1.10–1.34µm) 1700 4.5 2000 0.5 1700 4.0 1950 4.0 1500 4.0 1
HK(1.46–2.46µm) 1600 4.5 1500 3.5 1500 3.0 1650 4.7 1300 4.0 1
30-50MyrObjects
GSC08047B
JHK(1.10–2.46µm) 2200 3.5 2200 1.0 1900 4.5 2250 3.1 2300 4.0 2
J(1.10–1.34µm) 2600 5.0 2200 3.5 1600 3.5 2300 4.2 ≥2400b 4.0 2
HK(1.46–2.46µm) 1800 6.0 1700 4.5 1700 4.0 1850 5.5 1600 4.5 1
2MASS0141
JHK(1.10–2.46µm) 1800 5.0 2000 3.0 1700 4.0 2000 5.5 1600 4.5 1
J(1.10–1.34µm) 2300 4.0 2200 3.5 1600 3.5 2300 4.9 ≥2400b 4.0 2
HK(1.46–2.46µm) 1800 6.0 2100 1.0 1800 5.0 1900 5.5 1600 4.5 1
ABPicB
JHK(1.10–2.46µm) 1800 6.0 1600 3.5 1600 4.5 1850 5.5 1400 4.5 1
J(1.10–1.34µm) 2400 4.5 2300 4.0 1600 3.5 2400 4.9 ≥2400b 4.0 2
HK(1.46–2.46µm) 1900 6.0 2300 1.5 1800 5.0 1950 5.5 1700 5.0 1
Note: aThesestarformationregionmembershavebeendereddenedasdescribedinSect.4.bNotethatthisTeff fallsattheedgeofthegridused.
with the full grid summarized in Table 5 and then with a re- Cardellietal.(1989)usingtheextinctionmeasuredforthepri-
stricted subset with limits on the log(g) values appropriate to mary,asdescribedinSection4.
theagesofthetargets,basedonevolutionarymodels(Chabrier TheoverallspectralshapeisdominatedbytheH Oabsorp-
2
et al. 2000). Finally, the impact of metallicity is explored with tionatbothsidesofeachwavelengthbandandtheCOabsorp-
a comparison of the data with the expanded grid with different tion band at the red edge of the K-band. All the object spectra
metallicitiesfortheDrift-PHOENIXmodel. exhibit a distinct triangular shape in the H-band portion of the
spectrum, a signature of young substellar objects. As noted in
otherspectraofyoungobjects,thepeakoftheK-bandalsooc-
6. ResultsandDiscussion
cursatalongerwavelengthnear2.3µmthanistypicalforolder
6.1. SINFONISpectra fieldL-dwarfs(Mohantyetal.2007).Twoofthecoolestobjects,
2M1207BandABPicB,showanunusuallydepressedJ-band
The J,H,K flux-calibrated SINFONI spectra for the nine young portion of the spectrum relative to the spectra of the other sub-
substellarobjectsinthesampleareplottedinFigure5,grouped stellarobjectsofthesameage.
in the three age bands and ordered in decreasing temperature
determined from the model comparisons described in Section
6.2.ApproximatelyhalfoftheSINFONIdataarepreviouslyun- 6.2. ComparisonwithAtmosphereandEvolutionaryModels
published: all wavelength bands of DH Tau B, 2M1207B, and
6.2.1. FullModelGrid
2M0141 are newly reported along with the J- and H-bands of
TWA5BandGSC080647B.Fortheyoungesttargets,thespec- Eachobjectspectrumwascomparedwithagridofthefivesolar-
tra have been dereddened according to the reddening law of metallicitymodelssummarizedinSection5–DUSTY,BT-Settl,
8
Patience,King,DeRosa,Vigan,Witte,Rice,HellingandHauschildt:Near-IRspectraacrossthebrowndwarf/planetboundary
Table7.BestFittingAtmopshereModelParameters,RestrictedGridComparison
WavelengthRange DUSTY BT-Settl Drift-PHOENIX Gaia-Dusty Marleyetal.
Teff (K) logg Teff (K) logg Teff (K) logg Teff (K) logg Teff (K) logg fsed
1-3MyrObjects
DHTauBa
JHK(1.10–2.46µm) 2600 4.0 2400 3.0 2000 4.0 2350 3.7 ≥2400b 4.0 2
J(1.10–1.34µm) 2400 4.0 2200 3.5 1600 3.5 2350 4.0 ≥2400b 4.0 2
HK(1.46–2.46µm) 2600 3.5 2500 3.5 2700 4.0 2550 3.8 ≥2400b 4.0 2
GQLupBa
JHK(1.10–2.46µm) 1900 3.5 1900 3.5 1700 3.5 2050 3.5 1500 4.0 1
J(1.10–1.34µm) 2500 4.0 2400 3.5 2600 4.0 2500 4.0 ≥2400b 4.0 2
HK(1.46–2.46µm) 1700 4.0 1700 3.5 1700 4.0 1750 4.0 1500 4.0 1
CTChaBa
JHK(1.10–2.46µm) 1900 4.0 1600 4.0 1700 3.5 1950 4.0 1500 4.0 1
J(1.10–1.34µm) 1900 3.5 1400 4.0 1700 3.0 2000 3.9 1000 4.0 1
HK(1.46–2.46µm) 2500 3.5 2300 3.0 1800 4.0 2350 3.1 2400 4.0 2
8MyrObjects
2MASS1207A
JHK(1.10–2.46µm) 3200 4.0 2800 4.5 ≥3000b 4.5 2900 4.5 ≥2400b 4.5 1
J(1.10–1.34µm) 2800 4.5 2700 4.5 2900 4.5 2800 4.5 ≥2400b 4.0 2
HK(1.46–2.46µm) 3100 4.5 3000 3.0 ≥3000b 4.5 3050 4.5 ≥2400b 4.5 1
TWA5B
JHK(1.10–2.46µm) 2500 3.5 2300 3.0 2200 3.0 2450 3.3 ≥2400b 4.0 2
J(1.10–1.34µm) 2700 4.5 2600 4.0 2800 4.5 2700 4.5 ≥2400b 4.0 2
HK(1.46–2.46µm) 2500 3.5 2000 3.0 1700 4.5 2250 3.1 1600 4.5 1
2MASS1207B
JHK(1.10–2.46µm) 1600 3.5 1500 3.5 1500 4.0 1650 4.3 1300 4.0 1
J(1.10–1.34µm) 1700 4.5 2000 3.0 1700 4.0 1950 4.0 1500 4.0 1
HK(1.46–2.46µm) 1600 4.5 1500 3.5 1500 3.0 1650 4.5 1300 4.0 1
30-50MyrObjects
GSC08047B
JHK(1.10–2.46µm) 2200 3.5 2200 3.0 1900 4.5 2250 3.1 2300 4.0 2
J(1.10–1.34µm) 2600 5.0 2200 3.5 1600 3.5 2300 4.2 ≥2400b 4.0 2
HK(1.46–2.46µm) 1700 5.0 1700 4.5 1700 4.0 1750 5.0 1600 4.5 1
2MASS0141
JHK(1.10–2.46µm) 1800 5.0 2000 3.0 1700 4.0 2000 5.0 1600 4.5 1
J(1.10–1.34µm) 2300 4.0 2200 3.5 1600 3.5 2300 4.9 ≥2400b 4.0 2
HK(1.46–2.46µm) 1800 5.0 2000 3.0 1800 5.0 1850 5.0 1600 4.5 1
ABPicB
JHK(1.10–2.46µm) 1700 5.0 1600 3.5 1600 4.5 1800 4.8 1400 4.5 1
J(1.10–1.34µm) 2400 4.5 2300 4.0 1600 3.5 2400 4.9 ≥2400b 4.0 2
HK(1.46–2.46µm) 2500 3.5 2300 3.0 1800 5.0 2300 5.0 1700 5.0 1
Note: aThesestarformationregionmembershavebeendereddenedasdescribedinSect.4.bNotethatthisTeff fallsattheedgeofthegridused.
Drift-PHOENIX,Gaia-Dusty,andMarleyetal.Therangeinthe 6.2.2. Restrictedlog(g)ModelGrid
parameters of Teff and log(g) are summarized in Table 5. The
modelfitswereperformedoneachfullJ,H,KspectrumandtheJ Sincethetargetsallhavewell-constrainedyoungages,thespec-
andH+Ksubsetstoinvestigatethevariationofinferredphysical tra were also compared with a restricted model grid limited
propertiesonthewavelengthcoverageofthedata.Theresultsof to log(g) values appropriate for each age range, as detailed in
themodelfitsforTeff andlog(g)aregiveninTable6,dividedby Section4andbasedontheevolutionarymodelsshowninFigure
theagegroupsandorderedbyaverageTeff fromtheJ,H,Kfit. 2 (Chabrier et al. 2000). Further support for using a restricted
log(g) grid comes from measurements of the Na index defined
BasedontheJ,H,Kspectrumfits,theeffectivetemperatures in(Allersetal.2007)andshowntobesensitivetosurfacegrav-
range from 1100 K to 3100 K and log(g) of 1.0 to 6.0, consid- ity,butindependentofspectraltype.Forthesixtargetsforwhich
ering the extremes of any given model fit. Some of the values the J-band data had sufficient sensitivity (2M1207A, TWA 5B,
for log(g) are outside the bounds expected from evolutionary ABPicB,GSC08047B,and2M0141),fivehadindicesranging
models, and the fits also do not show a clear trend of increas- from 0.93 to 1.08, within the range associated with giant stars
ing log(g) with increasing age. The lack of a clear correlation and young brown dwarfs in star-forming regions (Allers et al.
between log(g) and age may simply be a consequence of the 2007).Only2M0141hasahighervalueof1.14,thoughthistar-
coarsenessofthegridsandthedominantimpactoftemperature getmaybeolderthantherestofthesample.Theinferredvalues
over surface gravity in shaping the overall spectral energy dis- of Teff from the restricted grid model fits to the data are used
tribution, as opposed to specific gravity-sensitive features. The to estimate the range of possible temperatures consistent with
range of masses included in the sample also complicates the theobservations.Uncertaintiesbasedoncomparisonsofthefive
comparisonoflog(g)andage. differentsyntheticatmospheresshouldbemorerepresentativeof
9
Patience,King,DeRosa,Vigan,Witte,Rice,HellingandHauschildt:Near-IRspectraacrossthebrowndwarf/planetboundary
DH Tau B GQ Lup B CT Cha B
DUSTY DUSTY 1.5 DUSTY
1.5 2600/4.0 1.5 1900/3.5 1900/4.0
1 1 1
0.5 0.5 0.5
0 0 0
BT(cid:239)Settl BT(cid:239)Settl 1.5 BT(cid:239)Settl
1.5 2400/3.0 1.5 1900/3.5 1600/4.0
1 1 1
0.5 0.5 0.5
0 0 0
DRIFT [M/H]=0 DRIFT [M/H]=0 1.5 DRIFT [M/H]=0
1.5 2000/4.0 1.5 1700/3.5 1700/3.5
(cid:104) 1 (cid:104) 1 (cid:104) 1
f f f
0.5 0.5 0.5
0 0 0
Gaia(cid:239)DUSTY Gaia(cid:239)DUSTY 1.5 Gaia(cid:239)DUSTY
1.5 2350/3.1 1.5 2050/3.0 1950/3.0
1 1 1
0.5 0.5 0.5
0 0 0
Marley et al. Marley et al. 1.5 Marley et al.
1.5 2400/4.0/2 1.5 1500/4.0/1 1500/4.0/1
1 1 1
0.5 0.5 0.5
0 0 0
1 1.2 1.4 1.6 1.8 2 2.2 2.4 1 1.2 1.4 1.6 1.8 2 2.2 2.4 1 1.2 1.4 1.6 1.8 2 2.2 2.4
Wavelength (µm) Wavelength (µm) Wavelength (µm)
Fig.6.TheSINFONIspectra(blacklines)andbest-fitmodel(coloredlines)tothefullspectrumisplottedforthestar-formingregiontargets:DH
TauB(left),GQLupB(middle),andCTChaB(right).Therestrictedlog(g)modelgridsareused.
2M1207 A TWA5 B 2M1207 B
1.5 D32U0S0T/4Y.0 2 D25U0S0T/3Y.5 1 D16U0S0T/3Y.5
1.5
1
1 0.5
0.5
0.5
0 0 0
1.5 B28T0(cid:239)0S/4e.t5tl 2 B23T0(cid:239)0S/3e.t0tl 1 B15T0(cid:239)0S/3e.t5tl
1.5
1
1 0.5
0.5
0.5
0 0 0
1.5 D30R0I0F/T4 .[5M/H]=0 2 D22R0I0F/T3 .[0M/H]=0 1 D15R0I0F/T4 .[0M/H]=0
1.5
(cid:104) 1 (cid:104) (cid:104)
f f 1 f 0.5
0.5
0.5
0 0 0
1.5 G29a0ia0(cid:239)/4D.5USTY 2 G24a5ia0(cid:239)/3D.0USTY 1 G16a5ia0(cid:239)/4D.1USTY
1.5
1
1 0.5
0.5
0.5
0 0 0
1.5 MOuatr loefy reatn agle. 2 M24a0r0le/y4 .e0t/ 2al. 1 M13a0r0le/y4 .e0t/ 1al.
1.5
1
1 0.5
0.5
0.5
0 0 0
1 1.2 1.4 1.6 1.8 2 2.2 2.4 1 1.2 1.4 1.6 1.8 2 2.2 2.4 1 1.2 1.4 1.6 1.8 2 2.2 2.4
Wavelength (µm) Wavelength (µm) Wavelength (µm)
Fig.7.TheSINFONIspectra(blacklines)andbest-fitmodel(coloredlines)tothefullspectrumisplottedfortheTWHydramembertargets:
2M1207A(left),TWA5B(middle),and2M1207B(right).Therestrictedlog(g)modelgridsareused.
thetruerangeofvaluesthanuncertaintiesestimatedfromacom- peak than the data. The peak of the H-band is also shaped by
parisonwithasinglemodelset. the dust cloud opacity, which is largely flat across the H-band.
A high cloud opacity can flatten the spectrum over this wave-
TheSINFONIdataandbest-fitatmospheremodel(restricted
lengthrange,andmayexplainthemodeldiscrepancyintheH-
grid)fromeachofthefivemodelsisshowninFigures6,7,and
band shape. The portion of the data with the lowest signal-to-
8forthethreeagegroups.Theregionofthespectrummostdiffi-
noiseistheblueedgeoftheJ-band,andimprovingthissection
culttofitforthemodelsisthepeakandslopeonthebluesideof
of the data would enable more reliable fits, as some of the tar-
the H-band. The models often show a steeper slope and higher
10