PDF (1.172 MB)

17 Pages·2014·1.17 MB·English

A&A564,A11(2014) Astronomy
DOI:10.1051/0004-6361/201220488 &
(cid:2)c ESO2014 Astrophysics
Origin of the wide-angle hot H in DG Tauri
2
New insight from SINFONI spectro-imaging
V.Agra-Amboage1,S.Cabrit2,4,C.Dougados3,4,L.E.Kristensen5,L.Ibgui2,andJ.Reunanen6
1 Universidade do Porto, Faculdade de Engenharia, Departamento Engenharia fisica, SIM Unidade FCT 4006, rua Dr. Roberto
Friass/n,4200-465Porto,Portugal
e-mail:[email protected]
2 LERMA,UMR8112duCNRS&ObservatoiredeParis,ENS,UPMC,UCP,61avenuedel’Observatoire,75014Paris,France
3 Laboratoire Franco-Chilien d’Astronomie, UMI 3386 du CNRS, 1515 Camino el observatorio, Casilla 36-D correo central,
Santiago,Chile
4 IPAG,UMR5274duCNRS&Univ.JosephFourier,414ruedelaPiscine,38041Grenoble,France
5 Harvard-SmithsonianCenterforAstrophysics,60GardenStreet,CambridgeMA02138,USA
6 TuorlaObservatory,DepartmentofPhysicsandAstronomy,UniversityofTurku,Väisäläntie20,21500Piikkiö,Finland
Received1October2012/Accepted30December2013
ABSTRACT
Context.Theoriginofprotostellarjetsremainsamajoropenquestioninstarformation.Magnetohydrodynamical(MHD)discwinds
areanimportantmechanismtoconsider,becausetheywouldhaveasignificantimpactonplanetformationandmigration.
Aims.WewishtotesttheoriginsproposedfortheextendedhotH at2000KaroundtheatomicjetfromtheTTauristarDGTau,in
2
ordertoconstrainthewide-anglewindstructureandthepossiblepresenceofanMHDdiscwindinthisprototypicalsource.
Methods.Wepresent spectro-imaging observations of theDGTaujetinH 1–0 S(1)with0(cid:3).(cid:3)12angular resolution, obtainedwith
2
SINFONI/VLT. Thanks to spatial deconvolution by the point spread function and to careful correction for wavelength calibration
andforunevenslitillumination(towithinafewkms−1),weperformedathoroughanalysisandmodeledthemorphologyandkine-
matics. Wealso compared our resultswith studiesin [FeII], [O I], and FUV-pumped H . Absolute flux calibration yields theH
2 2
column/volumedensityandemissionsurface,andnarrowsdownpossibleshockconditions.
Results.Thelimb-brightenedH 1–0S(1)emissioninthebluelobeisstrikinglysimilartoFUV-pumpedH imaged6yrlater,con-
2 2
firmingthattheytracethesamehotgasandsettinganupperlimit<12kms−1onanyexpansionpropermotion.Thewide-anglerims
areatlowerblueshifts(between–5and0kms−1)thanprobedbynarrowlong-slitspectra.Weconfirmthattheyextendtolargerangle
andtolowerspeedtheonion-likevelocitystructureobservedinopticalatomiclines.Thelatterisshowntobesteadyover≥4yrbut
undetectedin[FeII]bySINFONI,probablyduetostrongirondepletion.Therimthickness≤14AUrulesoutexcitationbyC-type
shocks,andJ-typeshockspeedsareconstrainedto(cid:6)10kms−1.
Conclusions. We find that explaining the H 1–0 S(1) wide-angle emission with a shocked layer requires either a recent out-
2
burst (15 yr) into a pre-existing ambient outflow or an excessive wind mass flux. A slow photoevaporative wind from the dense
irradiateddiscsurfaceandanMHDdiscwindheatedbyambipolardiffusionseemtobemorepromisingandneedtobemodeledin
moredetail.Betterobservationalconstraintsonpropermotionandrimthicknesswouldalsobecrucialforclarifyingtheoriginofthis
structure.
Keywords.stars:formation–stars:winds,outflows–stars:variables:TTauri,HerbigAe/Be–techniques:imagingspectroscopy–
moleculardata–infrared:stars
1. Introduction proximity and rich set of diagnostic emission lines, jets from
youngstellarobjectsrepresentaparticularlyusefultestbedfor
Jetsfromyoungstarsareubiquitousinstarformation,andtheir theMHDjet-launchingparadigm(seethereviewsofBallyetal.
link with accretion processes is well established (Cabrit et al. 2007;Rayetal.2007).
1990;Hartiganetal.1995).Theyarelaunchedfromtheinnerre- Molecular hydrogen is the primary gas constituent in the
gionsofthestar-discsystem,thesamediscthatmayultimately circumstellarenvironmentofyoungstars.Thev=1–0S(1)tran-
giverisetoaplanetarysystem.Althoughanumberofmodelsex- sitionofH at2.12μmisoftendetectedinembeddedprotostars,
2
istfortheirgeneration(see,forexample,thereviewsofFerreira withtheemissionformedprimarilyinshock-excitedcollimated
et al. 2006;Shanget al. 2007;and Pudritz et al. 2007),the ex- outflows(e.g., Davis et al. 2011,2002).The presence of H in
2
actmechanismbywhichmassisejectedfromaccretingsystems jets might be an indication that MHD ejection operates out to
and collimated into jets is an open problem. It is thought that disc radii≥0.2–1AU wheremoleculescansurviveagainstdis-
thesamebasicmagnetohydrodynamic(MHD)mechanismisre- sociation (Safier 1993;Panoglouet al. 2012).If such extended
sponsibleforlaunchingjetsinobjectsasdiverseasbrowndwarfs MHD disc winds persist until the later planet building phase,
(Whelan et al. 2005), post-AGB stars (e.g., García-Díaz et al. they could maintain fast accretion across the disc “dead-zone”
2008;García-Seguraetal.2005),compactobjectsinbinarysys- whereionizationinthemidplaneistoolowtoactivateMHDtur-
tems, and perhapseven active galactic nuclei. Because of their bulence(Bai2013).Theassociatedmagneticfieldswouldhave
ArticlepublishedbyEDPSciences A11,page1of17
A&A564,A11(2014)
strong consequences for the formation and migration of plan- TTS, namely: (1) an irradiated disc/envelope; (2) a molecular
ets(Fromangetal.2005;Bai &Stone2013).TheH emission MHD disc wind heated by ambipolar diffusion; (3) a forward
2
fromwarm(T =1000–3000K)gaswasrecentlydetectedwithin shocksweepinguptheenvelope;(4)areverseshockdrivenback
100AUofmoreevolvedTTauristars(TTSs)andHerbigAe/Be intoawide-anglemolecularwind.Thepaperisorganizedasfol-
stars, both in the near-infrared and far-ultraviolet (FUV) do- lows: The observations,data reductionprocedure,and centroid
mains (e.g., Valenti et al. 2000; Ardila et al. 2002; Bary et al. velocity measurements are described in Sect. 2. In Sect. 3, we
2003;Carmonaetal. 2011).Whenstudiedathighspectralres- presentthenewresultsbroughtbyourobservationsonthemor-
olution, these lines split into two populations: most show nar- phology,kinematics,andcolumndensityofthewide-angleH ,
2
rowprofilescenteredonthestellarvelocity,suggestinganorigin andcomparewiththeatomicandFUVH emissionproperties.
2
in a warm disc atmosphere (quiescent H ), but (cid:6)30% of them In Sect. 4, we clarify the shock parameters that could explain
2
display profiles blueshifted by 10–30 km s−1, sometimes with the H surface brightness and line ratio, we explore the possi-
2
bluewingsextendingup to –100kms−1, suggestingformation ble 3D velocity fields, and we discuss the strengths and weak-
in an outflow(Takamietal. 2004;Herczegetal. 2006;Greene nesses of various proposed origins for the wide-angle H . Our
2
etal.2010;Carmonaetal.2011;Franceetal.2012).Inpartic- conclusions are summarized in Sect. 5. Throughout the paper
ular, long-slit spectra of DG Tau provide evidence of a warm weassumeadistanceof140pctotheobject.
H outflow at (cid:6)–20 km s−1, shifted by (cid:6)30 AU along the blue
2
jetdirectionandwithanestimatedwidth(cid:6)80AU(Takamietal.
2. H observations,datareduction,andvelocity
2
2004; Schneider et al. 2013b). Beck et al. (2008) conducted a
measurements
2Dspectro-imagingstudyofH rovibrationallinesinsixCTTs
2
known to be associated with large scale atomic jets, including ObservationsoftheDGTaumicrojetinK bandwereconducted
DGTau,andfoundspatiallyextendedemissioninallcases,with on 15 October 2005 at the Very Large Telescope (VLT), using
sizes up to 200AU. While the data are notflux calibrated,rel- theintegralfieldspectrographSINFONIcombinedwithanadap-
ative levelpopulationsamong K-bandH2 linesindicate excita- tive optics (AO) module (Eisenhauer et al. 2003; Bonnet et al.
tiontemperaturesintherange1800–2300K.Fromvariousargu- 2004). The field of view is 3(cid:3)(cid:3)×3(cid:3)(cid:3) with a spatial sampling of
ments,Takamietal.(2004)andBecketal.(2008)arguethatthe 0(cid:3).(cid:3)05×0(cid:3).(cid:3)1per“spaxel”.Thesmallestsamplingisalongthedi-
H2propertiesareingeneralnotwellexplainedbyFUVorX-ray rection of the slicing mirrors(denoted as X-axis in the follow-
irradiationandaremostconsistentwithshock-excitedemission ing) and is aligned perpendicular to the DG Tau jet axis, i.e.,
from the inner regions of the atomic jets or from wider-angle at PA = 45◦. After AO correction, the effective spatial reso-
windsencompassingtheseflows.Takamietal.(2004)alsomen- lution achieved in the core of the point spread function (PSF)
tion ambipolar diffusion on the outer molecular streamlines of is ∼0(cid:3).(cid:3)12 (FWHM of a Gaussian fit to the reconstructed con-
anMHDdiscwindasanalternativeheatingmechanismforthe tinuumimage). The spectralconfigurationused providesa dis-
warmH2outflowinDGTau.Schneideretal.(2013a)alsofavor persion of 0.25 nm/pixel over a spectral range of 1.95 μm to
a molecular MHD disc wind, but heated by shocks rather than 2.45 μm. The spectral resolution as determined from OH sky
by ambipolar diffusion. In HL Tau, on the other hand, Takami lines is ∼88 km s−1 (median FWHM), with smooth systematic
et al. (2007)interpretthe V-shapedH2 emission encompassing variations of ±15 km s−1 along the Y-axis. Three datacubes of
the atomic jet as a shocked cavity driven into the envelope by 250seachwereobtained,oneonskyandtwoonsource(referred
anunseenwide-anglewind.Inanycase,extendedH2 emission toasexposure1andexposure2hereafter).
in TTSclearlyholdskeyinformationontheirwide-anglewind The standard data reduction steps were carried out using
structure, the radial extension of any MHD disc wind present, the SINFONI pipeline. Datacubes are corrected for bad pix-
anditsimpactonprotoplanetarydiscphysics. els, dark, flat field, geometric distortions on the detector, and
In an effort to extract this information, we present in this sky background. An absolute wavelength calibration based on
paper an in-depth analysis of H ro-vibrational1–0 S(1) emis- daytime arclamp exposuresis also performedby the SINFONI
2
sion associated with DG Tauri, using flux-calibrated spectro- pipeline. However, the accuracy is insufficient for kinematical
imaging K-band observations with 0(cid:3).(cid:3)12 resolution conducted studies of stellar jets. Unlike for the H-band data of Paper I,
withSINFONI/VLT.Thisworkcomplementsourdetailedstudy the OH sky lines in our K-band sky exposure have too low
of the atomic flow component and accretion rate in DG Tau signal-to-noise ratio (S/N) to provide a reliable improvement.
basedon[FeII]linesinH-bandandBrγinK-band,observedat Instead, we rely here on absorption lines of telluric and pho-
thesameepochwiththesameinstrument(Agra-Amboageetal. tosphericoriginagainstthe strongDG Taucontinuum.Telluric
2011,hereafterPaperI).OuranalysisimprovesonthatinBeck lines near the H 1–0 S(1) line were identified from compar-
2
et al. (2008) in several ways: We apply spatial deconvolution ison with a simulated atmospheric absorption spectrum con-
toobtainasharperviewofthebrightnessdistribution,andper- volved to our spectral resolution, generated with an optimized
formcarefulcorrectionforwavelengthcalibrationanduneven- line-by-linecodetestedagainstlaboratorymeasurements(Ibgui
slitilluminationtoretrieveradialvelocitiesalongandacrossthe &Hartmann2002;Ibguietal.2002),usingmoleculardatafrom
jet axisdownto a precisionof a few km s−1. This allowsus to the HITRAN database (Rothman et al. 2009). Two unblended
conducta detailedcomparisonwith the spatio-kinematicstruc- H O lines strong enough for velocity calibration were identi-
2
ture of the atomic flow component and FUV H emission in fied:oneat(cid:6)2.112μm(4735.229cm−1),andoneat(cid:6)2.128μm
2
DGTau,andtoconstrainboththepropermotionandthe3Dve- (4699.750cm−1). Three photospheric lines are also present on
locityfield of warmH , includingrotation.Inaddition,the ab- thebluesideoftheH 1–0S(1)line(Wallace&Hinkle1996):
2 2
solute flux-calibrationof the 1–0 S(1) line is used to infer col- a Mg doublet at (cid:6)2.107 μm (4746.841 and 4747.097 cm−1)
umndensitiesandemittingsurfacesand,togetherwithits ratio and two Al lines at (cid:6)2.110 μm and (cid:6)2.117 μm (4739.598and
to 2–1 S(1) at 2.25 μm, to restrict the possible shock parame- 4723.769cm−1).AllthreearewelldetectedintheDGTauspec-
tersforH excitation.Wethencombinethisnewinformationto trum (see also Fischer et al. 2011), but since the Mg line is a
2
critically re-examine the main scenarii that have been invoked doubletweonlyusethetwoAllinesforvelocitycalibration.In
for the origin of wide-angle H emission in DG Tau and other each exposure, the centroids of the two Al photospheric lines
2
A11,page2of17
V.Agra-Amboageetal.:Originofthewide-anglehotH inDGTauri
2
andtwoH OtelluriclinesweredeterminedbyGaussianfitting this effect depends on the gradient of light distribution within
2
onahighS/Nreferencespectrumobtainedbyaveragingspaxels the “slitlet”, and in principle can be as high as the spectral
over ±0(cid:3).(cid:3)5 along the slicing mirror closest to the DG Tau con- resolution. In our observations, the total light distribution is
tinuumpeak(centeredatY = 0inexposure1andY = 0(cid:3).(cid:3)03in dominatedby the PSF ofthe stellar continuum,whose FWHM
exposure2).After correctingphotosphericlines forthe motion isontheorderofthe “slitlet” width(0.1(cid:3)(cid:3))thereforeonemight
ofDG Tau(V = 16.5kms−1, Bacciottietal. 2000),thefour expectthiseffecttobesignificant.However,theAOsystemdoes
hel
velocityshiftswerefitted bya linearfunctionofwavelengthto notprovidefullcorrectionforatmosphericseeingonDGTau,so
derivetheglobalcalibrationcorrectiontoapplyat2.12μm.The thatmostof the fluxin the PSFis notin the Gaussian corebut
maximumdeviationbetweenthe4datapointsandthelinearfit in broader wings with smoother gradients. As we demonstrate
is ≤2km s−1 in each exposure,and we take thisas an estimate in Appendix A, the spurious velocity shifts in such conditions
ofthemaximumerrorinourabsolutecalibration. remain small (a few km s−1). Furthermore, they can be mod-
TheH 1–0S(1)emissionprofilesineachexposurearethen eled and corrected for with good accuracy, with an estimated
2
continuum-subtractedand corrected for telluric absorption fol- residual error of ≤2 km s−1, and even better in the blue lobe
lowing the procedure described in Paper I: We select a high beyond−0(cid:3).(cid:3)3 fromthe star. All velocitiesin thispaperare cor-
S/N reference“photospheric”spectrumwithno discernableH rected for uneven slit illumination using the model outlined in
2
line at our resolution (at Δx ∼ 0(cid:3)(cid:3), Δy ∼ 0(cid:3)(cid:3)). This reference Appendix A, and expressed in the reference frame of the star
spectrumisscaleddowntofitthelocalcontinuumlevelofeach assumingV =16.5kms−1(Bacciottietal.2000).
hel
individualspaxel, and subtracted out. This ensures optimal ac-
curacy in the removal of telluric and photospheric absorption
features(cf.Fig.1inPaperI);itwillalsoremovethePSFofany
3. Results
faint H emission presentin the reference position, but the ex-
2
tendedemissionbeyond0(cid:3).(cid:3)1ofthestarwillnotbeaffected.The 3.1.H 1–0S(1)morphologyandcomparisonwithFUV
2
same procedureis applied to extract the 2–1 S(1) line profiles. and[FeII]data
Correction for atmospheric differential diffraction is then done
byregisteringeachH datacubeonthe centroidpositionofthe ThetopleftpanelofFig.11showstherawcontinuum-subtracted
2
nearbycontinuum,ascalculatedby a Gaussian fit in the X and mapintheH21–0S(1)line2wherethetwoon-sourceexposures
Y directionsofthereconstructedcontinuumimage.Thedataare were combined to increase S/N. The morphology agrees with
flux-calibratedonthestandardstarHD28107,withanestimated thatfirstreportedbyBecketal.(2008)inthesamelineatcom-
errorof5%. parableresolution,namely:abrightcuspontheblueside(neg-
H centroidvelocitiesaredeterminedbyaGaussianfittothe ativeΔY),afainterbroaderarcontheredsideat+0(cid:3).(cid:3)6–1(cid:3)(cid:3)from
2
extracted line profile. The fitted centroids are subject to a ran- thestar,andadarklaneinbetween,presumablyduetoocculta-
domfittingerrorduetonoise,givenby1σ(cid:6)FWHM/(2.35S/N) tionbythecircumstellardiscofDGTau.Thedarklanewaspre-
foraGaussianofwidthFWHMandsignal-to-noiseratioS/Nat viouslyseeninopticalandH-bandobservationsoftheDGTau
thelinepeak(Porteretal.2004).Monte-Carlosimulationsshow jet, and the inferred disc size and extinction found consistent
that this error can be much smaller than the velocity sampling with disc properties from dust continuum maps (see Lavalley
(Agra-Amboage et al. 2009). The FWHM of fitted Gaussians etal.1997;Pyoetal.2003,andPaperI).
show the same systematic variations with Y-offset as OH sky Togiveasharperviewofthemorphology,therawimagewas
lines and indicate a small intrinsic width ΔV ≤ 30 km s−1 for deconvolvedbythecontinuumimage,usedasanestimateofthe
the H emission, in agreement with higher resolution spectra contemporarypointspreadfunction(PSF).We usedtheLUCY
(Takam2 ietal.2004;Schneideretal.2013b). restorationroutineimplementedintheSTSDAS/IRAFpackage.
The derived image did not change significantly after 40 itera-
After global absolute wavelength calibration, velocity cen-
tions.Theadvantageofdeconvolutionismainlytosuppressthe
troidsarestillsubjecttosmallresidualinstrumentalcalibration
driftsalong/acrossslicingmirrors.Oursimultaneousskyexpo- strongextendednon-GaussianwingsofthePSFcausedbypar-
sure has too low S/N to correct for this effect. We thus esti- tial AO correction. The final resolution is not precisely known
mate its typicalmagnitudeusing brightOH linesnear2.12μm but,givenourlimitedspatialsampling,isprobablyclosetoour
in two high S/N K-band sky datacubes obtained in the same GaussianPSFcorebeforedeconvolution(0(cid:3).(cid:3)12FWHM).
instrumental configuration, during a different run. Along each The deconvolvedimageof theblue lobeis presentedin the
slicingmirror,i.e.,transversetothejetaxis,thecentroidsshifts bottom left panel of Fig. 1. Bright H2 1–0 S(1) emission ap-
are mainly caused by a systematic residualslope, with random pearsconfinedtoasmallroundishregionofabout0(cid:3).(cid:3)5=75AU
fluctuations<1kms−1 ontopofit. Afteraveragingalongeach in diameter. The emission is clearly limb-brightened, suggest-
mirror(i.e.,whenbuildingaPVdiagramalongthejetaxis),the ing a “hollow cavity” geometry. It is brighter on the side fac-
OHcentroidsshowrandomdriftsfromrowtorowwithanrms ingthe star, andextendssidewaysupto ±0(cid:3).(cid:3)2 fromthe jet axis
of0.5–0.8kms−1.TheresultingoveralldistributionofOHline before closing back. A bright peak is also seen along the jet
centroidsamongindividualspaxelsinthefieldofviewisessen- axis at ΔY (cid:6) −0(cid:3).(cid:3)1, and a fainter one at ΔY (cid:6) −0(cid:3).(cid:3)4. Note
tiallyGaussian,withanrmsof1.5kms−1.We willincludethe thatwemaybeunderestimatingthefluxwithin<0(cid:3).(cid:3)1ofthestar
relevantrandomdispersioninourerrorbarsonH centroidve- due to our procedure for continuum and telluric removal (see
2
locities,andalsotakeintoaccountthesystematic“slopeeffect”
1 Figures1–3,6–10,B.1,andB.2areavailableincolourinelectronic
alongslicingmirrorswhensearchingfortransverserotationsig-
form.
natures(seeSect.3.2.3).
2 BecausethespectralresponseofSINFONIisnon-Gaussianandin-
Finally, the H velocity centroids need to be corrected for
2 cludes broad “shoulders” that may lie below thenoise level, weinte-
spuriousvelocityshiftsduetouneven-slitillumination.Indeed,
grated the line flux only over the three brightest velocity channels at
the slicingmirrorsin SINFONIbehavelike a long-slitspectro- –35.61,–0.61and+34.38kms−1.Afactorof1.4wasthenappliedin
graph where off-axis light suffers from a shift in wavelength ordertoaccountforthefluxlostinthesespectralwings(asestimated
on the detectorwith respectto on-axislight. The magnitudeof onthebrightestpositionsinourmap).
A11,page3of17
A&A564,A11(2014)
Fig.1.Topleft:rawcontinuum-subtractedmapofH 1–0S(1)lineemissioninDGTau.Contoursstartat2.7×10−4ergs−1cm−2sr−1andincrease
2
byfactorsof2.Thecrossdenotesthecentroidofthecontinuumimage(shownasaninsert).Bottomleft:H 1–0S(1)mapafterdeconvolutionby
2
thecontinuumimage.ThelackofH emissionatthecentralpositionresultsfromourcontinuumsubtractionprocedure.Rightpanel:deconvolved
2
H 1–0S(1)image(yellowcontours)superimposedonthedeconvolvedchannelmapsof[FeII]1.64μmobtainedatthesameepochandresolution
2
byAgra-Amboageetal.(2011),intwovelocityranges:MV(redcontours;–160kms−1 <V <120kms−1)andHV(backgroundcolourimage;
V <−160kms−1andV >120kms−1).
Sect. 2). With this limitation in mind, the morphology of our We willuse the peakbrightnessbeforedeconvolutionasa safe
deconvolved1–0S(1)imageisstrikinglysimilartotheFUVim- lower-limit.
ageofLyα-pumpedH obtainedinJuly2011bySchneideretal.
2
(2013a), and confirms their conclusion that the H2 wide-angle In the right-hand panel of Fig. 1, we compare the H2 1–0
emissioninDGTauappearsessentiallystationaryover6years. S(1) morphology with contemporaneous images from Paper I
We findthatthe brightH2 rimsnearthe base expandedby less of the atomic jet in [Fe II]1.64 μm in medium-velocity (MV:
than0(cid:3).(cid:3)1betweenthetwoepochs,implyingapropermotionex- –160 km s−1 < V < 120 km s−1) and high-velocity
pansion speed Vexp < 12 km s−1. Dedicated monitoring in the (HV:V <−160kms−1 and V > 120 km s−1) intervals. In the
sameinstrumentalsetupwouldbenecessarytogetamoreaccu- redlobe,H followswellthebaseofthebowshockfeatureseen
2
ratepropermotionvalue.Nevertheless,thecurrentlimitalready intheMVrangeofthe[FeII]emission,suggestingthatittraces
setstightconstraintsonscenarioswheretheH2 emissionwould thelow-velocitynon-dissociativewingsofthisbowshock.Inthe
originatefromtheinteractionofawide-anglewindwithcircum- bluelobe,in contrast,thebrightrimsofH emissionare much
2
stellargas(seeSect.4). broaderthanthe[FeII]jet,thattheylargelyencompass.Thisis
furtherquantifiedinthebottompanelofFig.2,whereweplotthe
InthetoppanelofFig.2weplotacutofthesurfacebright- transverseFWHMoftheH 1–0S(1)emissioninthebluelobe,
2
nessofH 1–0S(1)emissionalongthebluejetaxisasafunction as measured on the deconvolved image, as a function of dis-
2
of distance from the source, both before and after deconvolu- tancefromthesource.Thewidthat(cid:6)–0(cid:3).(cid:3)1isdominatedbythe
tionbythecontinuum.Theoverallfluxdecreasewithdistanceis bright inner peak, which is only marginally resolved transver-
verysimilartothatobservedinH FUVemission(seeFig.4of sally.Beyondthispoint,thewidthincreasesupto 0(cid:3).(cid:3)5ata dis-
2
Schneideretal. 2013a).Theratio of1–0S(1)to FUV H lines tanceof–0(cid:3).(cid:3)25,indicatinganapparenthalf-openingangle(cid:6)45◦
2
thus appears roughly constant with distance. The mean bright- similartothatdeterminedbySchneideretal.(2013a)fromtheir
nessleveltowardsthecavitycenterremainsaboutthesamebe- FUV image.This is muchwider than the 14◦ and 4◦ measured
fore and after deconvolution, (cid:6)3×10−3 ergs−1 cm−2 sr−1 and atthesame epochforthe MVandHV channelmapsof[FeII]
isthereforeveryrobust.Thepeakbrightnessismoreuncertain, (seeFig.2andPaperI).Atlargerdistances,theroundishshape
sincedeconvolutionhasabiastoenhancepeaksattheexpense oftheH emissionmakesitsFWHM decreaseagain,whilethe
2
ofextendedlower-levelemissionbeyond-0(cid:3).(cid:3)4 fromthesource. [FeII]jetkeepsitsnarrowopening.
A11,page4of17
V.Agra-Amboageetal.:Originofthewide-anglehotH inDGTauri
2
Fig.2. Top panel: cut of H 1–0 S(1) surface brightness along the
2
blue jet axis, before and after deconvolution by the PSF. The drop
influxinside–0(cid:3).(cid:3)1resultsfromour continuum subtractionprocedure.
Deconvoledpeakvaluesshouldbeviewedwithcaution.Bottompanel:
transverseFWHMofH2 1–0S(1)emissioninthebluelobeasafunc- Fig.3.ColourmapofcentroidvelocityoftheH21–0S(1)lineresulting
tionofthedistancetothestar,asmeasuredinthedeconvolvedimage. from a one-component Gaussian fit to each individual spectrum, and
Inred,wesuperimposetheFWHMof[FeII]1.64μmatthesameepoch correctedforuneven slitillumination.OnlyspaxelswithS/N > 5are
intwo velocityranges: MVB (–160 km s−1 < V < +50km s−1) and shown.Thecrossmarksthepositionofthestar.ContoursofS/N = 5,
HVB(V <−160kms−1),takenfromPaperI. 10, 15, 25, 35, 45aresuperimposed inblack ;thefittingerror dueto
noiseis1σ(cid:6)(40/S/N)kms−1(seetextandSect.2).
In contrast to the marked difference in width between H consistentwithpreviouslong-slitdatainthisregion.Thehighest
2
and [Fe II], Beck et al. (2008) noted that the V-shaped H blueshiftof(cid:6) −25±3kms−1 (atΔY = −0(cid:3).(cid:3)07andΔX < ±0(cid:3).(cid:3)1)
2
emission in DG Tau appears morphologically similar to the isinexcellentagreementwiththe−30±4kms−1observedatthe
low-velocity blueshifted [S II] emission mapped in 1999 with samepositioninFUVH linesina0(cid:3).(cid:3)2wideslit(Schneideretal.
2
HST by Bacciotti et al. (2000). (Schneider et al. 2013a) noted 2013b).SINFONIspectraintegratedoverthesameslitaperture
that the latter actually appears to fill-in the “voids” in the asTakamietal.(2004)alsoshowcentroidscompatiblewiththeir
limb-brightened FUV H emission, and concluded that the H reportedradialvelocityin1–0S(1)of−15±4kms−1 (namely:
2 2
“cavity” is not really hollow but filled by low-velocity atomic −8.5±3 km s−1 in exposure1 and −7±4 km s−1 in exposure
flow. A limitation to this argumentis that the [S II] HST maps 2, including errors in absolute calibration and uneven slit cor-
weretaken6–12yearsbeforetheH near-infraredandFUVim- rection). The only apparent difference with long-slit data is at
2
ages, respectively, while the crossing time at 50 km s−1 from ΔY (cid:6) −0(cid:3).(cid:3)3,whereourcentroidon-axis(cid:6) −5±4kms−1 issig-
the axis to the H rim is only 3 years. Data obtained closer in nificantlylowerthanthe−20±6kms−1 reportedbySchneider
2
timearethusnecessarytoconfirmthisconclusion.InSect.3.2.2 etal.(2013b).Thismaybeunderstoodbynotingthatthenarrow
wepresentsucha comparisonbothin spaceandvelocityusing 0(cid:3).(cid:3)2HSTslitisdominatedbyasmallaxialH knot(cf.theFUV
2
transverse long-slit [O I] data acquired by Coffey et al. (2007) imageofSchneideretal.(2013a)),whileourvelocitycentroids
only2yearsbeforeourSINFONIdata. aremeasuredontheraw(notdeconvolved)SINFONIdatacubes
with large non-Gaussian PSF wings, and are thus heavily con-
taminatedbythemuchbrighterH rimsatwiderangles.Theax-
2
3.2.H 1–0S(1)kinematics
2 ialH knotcouldalso betime variable.Overall,ourdata show
2
thatmostofthewide-angleH emissiontowardsthebluelobeis
3.2.1. 2Dvelocitymapand1Dvelocitygradientsalongthejet 2
atsignificantlylowerradialvelocitythaninferredfromlong-slit
Figure 3 presents a 2D map of the centroid velocities for ex- spectra.
posure2, computedas detailed in Sect. 2. Contoursof S/N are Inordertoincreaseprecisiononvelocitygradientsalongthe
superposedtoquantifythecentroidGaussianfittingerrorineach flowandintheredshiftedlobe,weincreasedS/Nbybuildingfor
spaxel.Thistermgenerallydominatesovertheuncertaintiesdue eachexposureaposition−velocity(PV)diagramalongthejetin
to instrumental drifts from spaxel to spaxel (1.5 km s−1 rms), apseudo-slitof1(cid:3)(cid:3)width,keepingtheoriginalY-samplingalong
uneven-slit illumination correction (≤2 km s−1), and absolute thejetaxisdefinedbythepositionsoftheSINFONIslicingmir-
calibration (<2 km s−1). Consistent results are found for expo- rors.Thefaintspectraatdistancesfartherthan–0(cid:3).(cid:3)7(blueside)
sure 1, considering the slightly different spaxel positions with andbetweenthestarand+0(cid:3).(cid:3)5(“darklane”ontheredside)were
respecttothecentralsource. furtheraveragedtoincreaseS/N.Wecarriedouttheanalysisin-
Figure3showsthatmostoftheareainthebluelobeisatlow dividuallyforeachexposure;sincetheircorrectionsforabsolute
radialvelocity,withatypicalvalue(cid:6)–5kms−1towardsthecav- calibrationandunevenslit-illuminationdiffer(seeSect.2),com-
itycenter.Thisisinlinewiththelowmedianvelocitybarycenter parisonoftheresultsgivesausefulcross-check,whilealsoim-
of–2.4kms−1 reportedbyBecketal.(2008)intheircompara- provingspatialsamplingofvelocitygradientsalongthejetaxis,
blespectro-imagingstudy.Atthesametime,wefindaspatially where the slicing mirror width (0(cid:3).(cid:3)1) undersamplesthe core of
limitedregionofhigherblueshiftsclosetothestarthatappears thePSF(0(cid:3).(cid:3)12FWHM).
A11,page5of17
A&A564,A11(2014)
Fig.4. Top: luminosity of H 1–0 S(1)
2
per 0(cid:3).(cid:3)1 length along the jet, integrated
over 1(cid:3)(cid:3) across the jet, for each expo-
sure. Bottom: Gaussian-fitted centroid
velocity of H 1–0 S(1) in spectra av-
2
eraged across a 1(cid:3)(cid:3) wide pseudo-slit.
Correction for uneven slit-illumination
was applied toeach averaged spectrum
(see Appendix A). Absolute velocity
calibration is better than 2 km s−1 in
each exposure. Vertical 1σ error bars
include the fitting error due to noise,
and 0.8 km s−1 rms instrumental cal-
ibration drifts between slicing mirrors
(see Sect. 2). Horizontal error bars in-
dicate where spectra werefurther aver-
agedalongthejettoincreaseS/N.
The top panel of Fig. 4 plots the luminosity inside each emissionin the velocityrangeofH 1–0S(1),andanapparent
2
“slice”ofthePVdiagram.ThebottompanelofFig.4plotsthe acceleration away from the star in opposite sense to the cen-
H centroidvelocitiesfittedtospectraaveragedacrossthejetas troidgradientobservedinH .Togetherwiththemuchnarrower
2 2
afunctionofdistancealongthejet,foreachexposureseparately. openingangleof[FeII]comparedtoH 1–0S(1)(cf.Sect.3.1)
2
Thetwo individualexposuresshow excellentagreementwithin thisappearstosuggestnodirectdynamicalinteractionbetween
the error bars, giving us confidence in our velocity determina- the[FeII]jetandthewide-angleH inDGTau.Thesamekine-
2
tions and error estimates. The velocity gradient along the blue maticandspatialseparationbetween[FeII]jetandH widean-
2
lobeseen in Fig. 3 is confirmed:the 1D-averagedH centroids gleemissionwasobservedpreviouslyinthecaseofHLTauby
2
showanapparentdecelerationfrom(cid:6)−18kms−1atΔY (cid:6)−0(cid:3).(cid:3)1 Takamietal.(2007),whotheninterpretedtheV-shapedH asa
2
to(cid:6)−5kms−1at−0(cid:3).(cid:3)2,andtoessentiallyzeronearthecavityend shockedoutflowcavityswept-upbyanunseenwide-anglewind.
atΔY =−0(cid:3).(cid:3)5.Thisgradientisclearlytoostrongtobecausedby In the case of DG Tau, previous optical spectro-imaging
measurementuncertainties;residualerrorsinunevenslitillumi- in [O I] and [S II] do provide direct evidence for a wide-
nationcorrectionareatmost2kms−1within-0(cid:3).(cid:3)3fromthestar, angle atomic wind, with a decreasing velocity from axis to
andappearnegligiblefurtherout(cf.AppendixA),whileresid- edge(Lavalley-Fouquetetal.2000;Bacciottietal.2000,2002).
ual errors in absolute calibration would only shift each curve However, these optical data were taken 6–7 years before the
withoutaffectingthegradients. SINFONIH data,atatimewhentheDGTaujetexhibitedvery
2
Inthefainterregionbeyondthecavityedge,bothexposures high-velocityemissionupto–350kms−1thatwasnotobserved
showatrendforanincreasedblueshiftat(cid:6)−10kms−1 atΔY (cid:6) since then. This makes a comparisonbetween the two datasets
−0(cid:3).(cid:3)6, that might be tentative evidence for a more distant faint somewhatuncertain.Inordertoinvestigatethespatio-kinematic
H2 knotalong the jet axis. This position correspondsto the tip relationshipbetweenthewide-angleH2 andtheatomicflowus-
of the bright [Fe II] HVC jet beam in Fig. 1. Even fainter H2 ing data as close in time as possible, we comparein Fig. 5 the
emission is detected around–1(cid:3)(cid:3), with centroidvelocities close transverse velocity structure at ΔY = −0(cid:3).(cid:3)3 across the blue jet
tozero.A faintbubblefeatureisseenatthispositionin[FeII] in the 1–0 S(1) line (from this paper), in the [Fe II] 1.64 μm
HV and MV (see Fig. 1 and Paper I). Finally, the H2 emission line observed simultaneously (from Paper I) and in the optical
in the red lobe appears mostly redshifted, with a mean radial [OI]λ6300Ålineobservedonly2yearsearlierwithHST/STIS
velocity(cid:6)+5 km s−1. In the rest of this paper,we focuson the (from Coffey et al. 2007). The [O I] profiles were fitted by
bright wide-angle H2 cavity in the blue lobe, as the remaining two Gaussians by Coffey et al. (2007): a “high-velocity com-
H2featureshavetoolowS/Nformoredetailedanalysis. ponent” (HVC) and a “low-velocity component” (LVC). Both
centroidsarereportedinthefigure.The[FeII]profilesarenar-
rowerandunder-resolvedbySINFONI.Theywerethusfittedby
3.2.2. Comparisonwithkinematicsinatomiclines
singleGaussians,dominatedherebytheHVC.
The velocity range of [Fe II]1.64 μm emission in the blue Figure5showsthatwhilethe[OI]HVCcoincidesinveloc-
lobe of DG Tau appears totally distinct from that of H . The ityandspatialwidthwiththenarrow[FeII]HVC,the[OI]LVC
2
[FeII]1.64μm SINFONIspectraintheregionoftheH cavity is much slower and spatially broader; its radial velocity drops
2
have centroidsranging from –150 to –200 km s−1 (Paper I). A from–60kms−1 on-axisto–25kms−1 at±0(cid:3).(cid:3)25fromtheaxis,
higherresolutionlong-slit[FeII]1.64μm spectrumobtainedin andthespatialFWHMis(cid:6) 0(cid:3).(cid:3)2(Coffeyetal.2008).Therefore,
2001byPyoetal.(2003)revealsaseparatelower-velocitycom- the [O I] LVC is filling-in the “gap” between [Fe II] and H
2
ponentpeaked around –50 to –100 km s−1, but still very weak emission,bothspatiallyandkinematically.Inaddition,wenote
A11,page6of17
V.Agra-Amboageetal.:Originofthewide-anglehotH inDGTauri
2
Fig.5. Centroid velocity versus distance from the jet axis for H
2
1–0S(1)(filledsquares,thispaper),[FeII]1.64μm(stars,fromPaperI) Fig.6. Velocity shifts in H2 1–0 S(1) between symmetric positions
and [OI]λ6300 Å HVC and LVC (open symbols, from Coffey et al. on either side of the jet axis, after averaging across 3 slicing mirrors
2007), all measured along a 0(cid:3).(cid:3)1-wide transverse cut at ΔY = −0(cid:3).(cid:3)3 with−0(cid:3).(cid:3)4 ≤ ΔY ≤ −0(cid:3).(cid:3)2. The 1σ error bars include the fittingerror
fromthestarintothebluelobe.PositiveΔXistotheSE. due to noise and the 1 km s−1 rms instrumental drifts from spaxel to
spaxel. Bluelinesillustratethetypical rangeof spurious instrumental
velocity gradients along slicing mirrors, as measured from OH lines
thatthecentroidvariationsandspatialFWHMofthe[OI]LVC (seeSect.2).Residualerrorsinabsolutecalibrationandcorrectionfor
observed in 2003 by Coffey et al. (2008) are identical to mea- uneven-slitilluminationcancelouthere.
surementsobtained4yearsearlieratthesamedistancefromthe
star (see Fig. 1 in Bacciotti et al. 2002, and Fig. 14 in Maurri
etal. 2013).We thusfind thatthe spatio-kinematicstructureof
theLVCremainsremarkablystableovermorethanatransverse tosearchfortransversevelocitygradientspossiblyindicativeof
crossing time, despite strong variability of the HVC at veloci- rotationinthewide-angleH2lobe.
ties≥300kms−1.Thisfindingdefinitelystrengthensconclusions ThedifferencesincentroidvelocityofH 1–0S(1)between
2
basedonearlier1999HSTdatathatthewide-angleH2cavityin symmetricpositionsacrossthejetaxisarerepresentedinFig.6.
DGTauappearsasaslower,outerextensionofthe“onion-like” Notethatanyresidualerrorsinabsolutecalibrationanduneven-
velocitystructureoftheatomicflow(Takamietal. 2004;Beck slit illuminationcorrectioncancelout(the illuminationis sym-
etal.2008;Schneideretal.2013a). metricwithrespecttothejetaxis,tofirstorder).ToincreaseS/N,
Another important implication of this comparison is that we averagedthe 3 slicing mirrorscentered at distances ΔY be-
[Fe II] emission does not appear as a good tracer of the wide tween(cid:6)0(cid:3).(cid:3)2and(cid:6)0(cid:3).(cid:3)4fromthestar,whereH emissionisbroad-
2
angle atomic flow slower than 50 kms−1. While Pyo et al. de- est.WethendeterminedthevelocitycentroidbyGaussianfitting
tect [Fe II] LVC emission within ±0.15(cid:3)(cid:3) of the jet axis at ve- at each offset ΔX across the axis (keeping the original spaxel
locitiesbluerthan–50kms−1,thecomparisonoftransversePV sampling of 0(cid:3).(cid:3)05 in this direction). In exposure 2, spaxels are
diagrams in [Fe II] and [O I] (see Fig. 5 of Paper I) shows no notpositionedsymmetricallyoneithersideofthejetaxis,there-
detectable contribution in [Fe II] from the wider angle, slower fore centroid velocities at ΔX < 0 were lineary interpolatedat
gas below –50 km s−1 that so clearly stands out in [O I] (and positionssymmetricfromthoseatΔX >0beforecomputingve-
[SII]).Thisdifferencecannotbeexplainedbygradientsinden- locitydifferences.Individualresultsfromthetwoexposuresare
sityortemperature:the[FeII]1.64μmlinehasacriticaldensity shown in Fig. 6. Blue lines superposed in Fig. 6 illustrate the
intermediatebetweenthatof[O I]and[SII],anda similar up- expected range of spurious velocity shifts due to instrumental
perlevelenergy.Weproposethatitisdueinsteadtoironbeing effectsalongslicingmirrors(“slopeeffects”,seeSect.2).
increasingly depleted at lower speeds. Indeed, modeling of the When taken individually, all velocity differences plotted in
[Fe II] to [O I] line ratio in Paper I indicates a higher average Fig. 6 are compatible with zero to within their respective error
gas-phase iron depletion of a factor 10 in the interval –100 to bars,exceptinexposure1at±0(cid:3).(cid:3)1fromtheaxis.However,there
+10kms−1,comparedtoonlyafactor3intheinterval–250to isasystematicconsistenttrendatmostpositionsandinbothex-
–100 km s−1. Depletion could be so high in the velocity range posures for spectra at positive ΔX (SE) to be more blueshifted
–50 to –25 km s−1 that the correspondingwide-angleemission thanthoseatnegativeΔX(NW).Thiswouldcorrespondtorota-
seen in [O I] and [S II] remains undetectable in [Fe II] at the tioninthesamesenseastheCOdiscandopticaljetinDGTau
sensitivityandresolutionofSINFONI. (Testi et al. 2002; Bacciotti et al. 2002; Coffey et al. 2007).
Unfortunately, the observed trends are also of the same order
as the maximum residual instrumentalslope along slicing mir-
3.2.3. ConstraintsonH rotation
2 rors observed in OH sky lines (blue lines in Fig. 6). Although
The LVC centroids in optical forbidden lines are known to (i)mostinstrumentalresidualslopesareshallowerthanthis;(ii)
display clear asymmetries between opposite sides of the jet averagingover3slicingmirrorsshoulddecreasetheeffect;and
axis,with amplitudeandsignconsistentwith arotating,steady (iii)thereisonlya25%probabilitythatitwouldgointhesame
magneto-centrifugaldiscwindlaunchedoutto(cid:6)3AU(Bacciotti senseastheobservedgradientinbothexposures,aninstrumen-
etal.2002;Andersonetal.2003;Pesentietal.2004;Coffeyetal. talorigincannotbetotallyexcluded.SincewelackabrightOH
2007,cf. Fig. 5). We thereforemake use of ourSINFONI data skyexposuresimultaneouswithourdatatocorrectfortheeffect
A11,page7of17
A&A564,A11(2014)
spaxel-by-spaxel, we can only put an upper limit on the true region is L1−0S(1) = 1.9 × 10−5 L(cid:11), only 50% more than in
transversevelocityshiftsinH presentinourdata. the 0(cid:3).(cid:3)3-wide slit of Takami et al. (2004). The total mass of
2
To obtain a conservative upper limit on the possible rota- hot H2 at 2000 K, assuming LTE, is then 3× 10−8 M(cid:11), while
tionalshiftatΔX ±0(cid:3).(cid:3)2fromtheaxis(wheretheH rimspeak, the characteristic emitting area at average brightness level is
2
cf. Fig. 1), we take the average shift of the two exposures at A(cid:6) L1−0S(1)/(4π(cid:9)F1−0S(1)(cid:10))=2×1030cm2 (cid:6)(90AU)2.
this offset (–1.3 km s−1), add its 3σ error bar (–4 km s−1), We note that we do not include any correction for dust at-
and correct by the maximum expected instrumental effect in tenuation of the H 1–0 S(1) line fluxes. Although Beck et al.
2
theoppositedirection(cid:6)2kms−1 (sincethissituationcannotbe (2008)derivedaratherhighA =11.2±10magtowardstheH
V 2
excludedeither).We obtainan upperlimit ofV ≤ 7 km s−1 peak,theuncertaintyintheirestimateislarge,duetostrongtel-
shift
in the sense of disc rotation. In the ideal situation of a sin- luricabsorptioninH Q-branchlinesnear2.4μm.Furthermore,
2
gle flow streamline observed with infinite angular resolution, theopticaljetbrightnesskeepsincreasingallthewayupto0(cid:3).(cid:3)1
this would set an upper limit on the H flow rotation speed of fromthe star (Bacciottiet al.2000),suggestingthattheatomic
2
Vφ =Vshift/(2sini)≤5kms−1forajetinclinationtothelineof jet suffers only moderate circumstellar dust extinction. Hence,
sightof40◦.Inreality,convolutionbythePSFandcontribution any 11 mag obscuring layer would have to lie behind the blue
from several streamlines along the line of sight both act to re- atomic jet lobe yet in front of the blueshifted H peak, which
2
ducetheobservablevelocityshifts(Pesentietal.2004),andthe seemsverycontrived.Finally,thestrikinglysimilarmorphology
trueVφcouldbesomewhathigherthanthis. ofFUVH2 emissionimagedbySchneideretal.(2013a)tothat
seeninthenear-infrared1–0S(1)linealsoarguesstronglyfora
low A to the hot H . We thus assume that dust obscurationis
V 2
3.3.H temperature,columnandvolumedensities, negligibleinK-bandlines.
2
andemittingarea
Becketal.(2008)computedtheratiosofvariousro-vibrational 4. Originofthewide-angleH2 emissioninDGTau
lines of H towards the H peak position, and found that they
2 2 In this section, we re-examine possible origins for the wide-
werecompatiblewithathermalizedpopulationatT (cid:6) 2000K.
angle, limb-brightened H 1–0 S(1) emission in the blue lobe
2
To search for possible temperaturegradients,we computedthe
of DG Tau, taking into account the new information extracted
2–1S(1)/1–0S(0)ratioatseveralpositions,afteraveragingover
from our SINFONI data on the geometry, proper motion, ra-
0(cid:3).(cid:3)2 along the jet and 0(cid:3).(cid:3)5 across the jet to increase S/N in the
dial velocities, and surface brightnesses (cf. previous section).
fainter v = 2–1 line. We found a similar ratio at ΔY = –0(cid:3).(cid:3)2
In Sect. 4.1, we determine the allowed shock parameter space
and –0(cid:3).(cid:3)4 along the blue jet, of 0.07 ± 0.02 and 0.09 ± 0.02
able to explain both the 1–0 S(1) surface brightness and the
respectively, consistent with the 0.07± 0.03 reported by Beck 2–1S(1)/1–0S(1)lineratio.InSect.4.2,wemodelgeometrical
etal.(2008)towardstheH2 peak.Henceweassumea uniform andprojectioneffectsto constrainthevelocityfieldcompatible
temperatureT (cid:6) 2000Kinthefollowing.Wenotethatthe1–0
with the observed radial velocities. Finally, we combine these
S(0)/1–0S(1)ratio(cid:6)0.25±0.03observedtowardstheH peak
2 two sets of constraintsto examine the plausibility and require-
(Takamietal.2004;Becketal.2008;Greeneetal.2010)com-
mentsofseveralscenariosproposedfortheoriginofwide-angle
bined with Eq. (5) of Kristensen et al. (2007) indicates an H
2 H inDGTauandsimilarsources(e.g.,HLTau):
ortho/para ratio (cid:6)3 consistent with the LTE value. We assume 2
thesameholdsthroughouttheregion. – anirradiateddiscatmosphere/envelope;
Our flux-calibrated data in H2 1–0 S(1) can be converted – amolecularMHDdiscwindheatedbyambipolardiffusion;
into column densities in the upper level of the transition – aforwardshockdrivenintothedisc/envelope;
(v = 1, J = 3) through N1,3 = F1−0S(1) × (4π)/(hνulAul) = – a reverse shock driven back into a molecular wide-angle
3.8×1019 cm−2 ×F1−0S(1)(ergs−1cm−2sr−1).AssumingLTEat wind.
2000Kandanortho/pararatioof3,thetotalcolumndensityof
hot H2 gas is then given by NH2 = N1,3/1.28× 10−2 (Takami 4.1.Allowedshockparameterspace
et al. 2004). The inferred column density of hot H reaches
2
b4r×igh1t0n1e9scsm(cid:6)−02.0t1ow3aerrdgss−th1ecHm2−2psera−k1;actf-.0F(cid:3).(cid:3)1igf.r2o)m, wthheilsetathre(pteyapk- Bpleacnkaretshalo.c(k20m08o)declosmfproamredStmheithH2(1li9n9e5r)atwioitshinprDeGshoTcaku dweinth-
ical value at the center of the blue lobe is (cid:6)1019 cm−2 (central sity 2×106cm−3. They mention good agreement with both an
brightness (cid:9)F1−0S(1)(cid:10) (cid:6) 3×10−3 ergs−1cm−2 sr−1; cf. Fig. 2). 8 km s−1 non-dissociativeJ(ump)-typeshock,and a 35 km s−1
Sincetheinnerpeakmighthaveacontributionfromanaxialjet multifluidC(ontinuous)-typeshock(whereionsandneutralsare
knot (see FUV image of Schneider et al. 2013a), and the rims decoupled). We here use the additional constraint provided by
areclearlylimb-brightened,weconsiderthebrightnesstowards the calibrated 1–0 S(1) H surface brightness, together with
2
thebluelobecenterasourbestestimatefortheintrinsic“face- a larger grid of planar shocks computed by Kristensen et al.
on” column density of the layer producing the wide-angle H2 (2008),to exploremore extensivelythe shock parameter space
emission. Both our image and that of Schneider et al. (2013a) thatcouldreproducetheDGTauobservations.
suggest that the rims are thinner than 0(cid:3).(cid:3)1 = 14 AU, yieldinga The predicted H surface brightnesses depend on three pa-
2
olofwneHr2l≥im5it×to1t0h4ecvmo−lu3m. edensityofhotH2intheemittinglayer rdaemnseitteyr,s:ntHh,ea(cid:2)snhdoctkhevedliomcietnysViosn,ltehsespmreasghnoectkichyfidelrdogpeanranmucelteeur3s
The top panels of Fig. 4 plot the 1–0 S(1) luminosity per b ≡ B(μG)/ n (cm−3), where B is the magnetic field com-
0(cid:3).(cid:3)1lengthalongthejet,integratedoverafullwidthof1(cid:3)(cid:3)across H
ponentparalleltotheshockfront.Thegridincludesbothsingle-
thejet,foreachindividualexposure.EventhoughtheH2 emis- fluid J-type shocks with b = 0.1, and steady multifluid C-type
sion is narrowest at the base and widens out, the luminos-
ity still peaks at –0(cid:3).(cid:3)1 and then decreases gradually with dis- 3 This parameter is related to the Alfvén speed in the preshock gas
tance. The total luminosity integrated over the whole emitting throughV =b×1.88kms−1.
A
A11,page8of17
V.Agra-Amboageetal.:Originofthewide-anglehotH inDGTauri
2
shockswithbrangingfrom0.5to10.Preshockdensitiesrange
from104to107cm−3.Thepreshockgasisassumedfullymolec-
ularwithnoUVfieldandastandardISMdustcontentandcos-
micrayflux;thelatterdeterminesinparticulartheinitialabun-
dancesofchargedspeciesandatomicH.
We note that model predictions assume a face-on view. A
sideways view would increase the observed surface brightness
by a factor 1/cos(i) for an infinite slab. This effectdoes notin-
troduce large errors when fitting H emission near the head of
2
abowshock,wherethepath-lengthislimitedbythestrongcur-
vature (Gustafsson et al. 2010), but will be more severe for a
tiltedcavitywherethefrontsideisalmostentirelytangenttothe
line of sight, as suggestedhere by the brighterrim towardsthe
star(seeSect.4.2).Hence,wewillrathercomparetheface-one
shockpredictionswith the 1–0S(1)surface brightnesstowards
the centerofthe H cavity,wherelimb-brighteningeffectswill
2
beminimal.
InFig.7wecomparetheobserved1–0S(1)surfacebright-
ness and 2–1 S(1)/1–0 S(1) ratio with model predictions for
J-shocks with b = 0.1 (top panel) and C-shocks with b = 1
(bottompanel).Itmaybeseenthatthe2–1S(1)/1–0S(1)ratiois
muchmoresensitivetoV thanton ,withchangesofafactor2
s H
intheformerhavingastrongereffectontheratiothanachange
ofafactor10inthelatter.InthecaseofJ-shocks,thedatapoint
towards the cavity center is fitted with a high n (cid:6) 106 cm−3
H
andalowshockspeedV (cid:6)9–10kms−1.InthecaseofC-shocks
s
withb = 1,themodelspointtoalowerdensityn (cid:6) 104 cm−3
H
and a higher shock speed V (cid:6)52 km s−1. Similar plots for C-
s
shockswithb = 0.5andb = 10areshowninAppendixB.The
datapointfortheinnerpeakat–0(cid:3).(cid:3)1fromthestar(toptriangu-
lar symbol in Fig. 7) would require a roughly 10 times higher
preshock density than the central point. However the required
Fig.7. H surface brightness in the 1–0 S(1) line against the
shock speed would change by less than a factor 2. Therefore, 2
v=2−1S(1)/v = 1–0 S(1)ratioasobserved intheDGTaubluelobe
our conclusionson V are robust even if the surface brightness
s (lowertriangle:centerofbluelobe,uppertriangle:innerpeak)andas
atthe cavitycenterisaffectedby residualgeometricprojection predictedforplanarshockmodelsviewedface-onfromKristensenetal.
effects. (2008) (colour curves). Theerror barsonobserved surfacebrightness
Table 1 summarizesthe pairsof values(nH, Vs)compatible aresmallerthanthesymbolsize,whilethedifferencebetweenthetwo
with observations for each magnetic field parameter b. It may symbols illustrates the maximum uncertainty due to projection/limb-
beseenthatalargermagneticparameterbrequireshighershock brightening effects. Top panel: planar J-shocks with b = 0.1. Bottom
speedandlowerdensitytomatchtheobservations.Itisnotewor- panel: C-shocks with b = 1. Each model curve corresponds to adif-
ferent value of the preshock hydrogen nucleus density n , increasing
thythatinallcases,theallowedrangeofshockspeedreproduc- H
ingtheobserved2–1S(1)/1–0S(1)ratioisverynarrow,setting from bottom to top. The shock speed Vs increases to the right and
somevaluesaremarkedalongthecurvestoguidetheeye.Similarplots
strongconstraintsonscenariosofshock-heating.Foreachplanar for C-shocks with b = 0.5 and b = 10 are shown in Appendix B.
shockmodelabletoreproducetheH 1–0S(1)brightnessand2–
2 Note the narrow range of shock speeds reproducing the observed
1/1–0ratio,wealsocalculatedVn(H2),the(emission-weighted) 2–1S(1)/1−0S(1)ratios.
centroidvelocityoftheH 1–0S(1)emissionlayerintherefer-
2
enceframeoftheshockwave.InJ-shocks,wheredecelerationis
almostinstantaneous,Vn(H2)isessentiallyzero.Incontrast,we alongthebluejetaxis.Theyareassumedinfinitelythinandwith
findthatitistypically(cid:6)60%oftheshockspeedintheC-shocks uniformsurfacebrightness.Webuiltsynthetic3Ddatacubesfor
ofTable1,wheredecelerationoftheneutralsoccursthroughcol- severalinclinationstothelineofsight.Eachchannelmapiscon-
lisionswiththe(rare)ionsandisthereforemuchmoregradual. volvedbya GaussianPSFofFWHM 0(cid:3).(cid:3)12,andthePSFofthe
spectrumat(0,0)issubtractedout(asperformedinourobserva-
tionsforthe purposeoftelluricandcontinuumsubtraction,see
4.2.Constraintsontheintrinsicvelocityfield
Sect.2).Residualemissionmapsarethenbuiltandcomparedby
ofthewide-angleH
2 eyewiththedeconvolvedimageofFig.1,todeterminethege-
Thelimb-brightenedmorphologyofthewide-angleH emission ometricalparametersreproducingtheH cavityshape,size,and
2 2
issuggestiveofatiltedcavity.Hence,observedradialvelocities limb-brightenedappearance.
areaffectedbycomplexprojectioneffects.Wethusneed3Dgeo- Inthecaseofcones,thedeconvolvedmapisbestreproduced
metricalmodelstoinferthetruevelocityfieldoftheH emitting with a sphericalradiusR (cid:6) 0(cid:3).(cid:3)43 (60 AU) and half-opening
2 max
gasparallel(V(cid:13))andperpendicular(V⊥)tothecavitysurface.To angle θ = 30◦ ± 3◦ for inclinations i = 45◦±5◦ to the line
remain as generalas possible, we consider two basic toy mod- of sight. A typical predicted map for θ = 27◦ and i = 40◦ is
els: (1) a hollow cone with its apex at the source position and shown in the top panel of Fig. 8, superposed on the observed
its axis aligned with the blue jet; and (2) a (half or full) hol- deconvolvedmap.Thefrontsideofthecavityisalmosttangent
lowspherecenteredataprojecteddistancez fromthesource tothelineofsight,sothatthelimb-brightenedrimgetsbrighter
proj
A11,page9of17
A&A564,A11(2014)
Table1.PlanarshockparametersforthepeakandcentralpositionsintheH cavity,assumingaface-onview.
2
(cid:2) J-shock C-shocks
b=B(μG)/ n (cm−3) 0.1 0.5 1.0 2.0 3.0 4.0 5.0 10.0
H
n (cm−3) 107 105−106 5×104−105 ∼4×104 ∼3×104 ∼3×104 ∼2×104 104
Peak H
V (kms−1) 8 32–16 41–35 ∼50 ∼59 ∼65 ∼70 88
s
n (cm−3) 106 (cid:6)5×104 (cid:6)104 ≤104 ≤104 ≤104 ≤104 <104
Center H
V (kms−1) 9–10 (cid:6)30 52 ≥60 ≥65 ≥70 ≥75 >90
s
closertothestar,asobserved;thenarrowpeakobservedat–0(cid:3).(cid:3)1
isnotfullyreproduced,andmayincludecontributionfromajet
knot.Theconesurfaceis(75AU)2,ingoodagreementwiththe
fullemissionsurface(cid:6)(90AU)2estimatedinSect.3.3.
In the case of a sphere, the best fit to the overall size and
shapeisforaradiusR (cid:6)0(cid:3).(cid:3)2(28AU)centeredatz (cid:6)0(cid:3).(cid:3)28
max proj
(40AU)alongthebluejet.However,toobtainabrighterrimat
thebase,theemissivitymustbestronglyenhancedinthehemi-
spherefacingthestar.AfterconvolutionbyourPSF,themapand
centroid velocities are then undistinguishable from those pre-
dicted for that single hemisphere. For simplicity, we thus con-
siderahalf-sphereinthefollowing.Thepredictedmapiscom-
paredwith theobservedoneinthe bottompanelofFig.8.The
emissionsurfaceof(70AU)2isclosetothatoftheconemodel.
For each geometry that reasonably reproduces the H 1–0
2
S(1)image,wethenconstructasyntheticPVdiagramalongthe
jetaxis,afterconvolvingthedatacubespatiallybyaMoffatfunc-
tion(agoodfittotherawPSFdeliveredbyourAOsystem)and
spectrally by a Gaussian of FWHM 88 kms−1 (median resolu-
tion of SINFONI overthe field of view).We then computeve-
locitycentroidsandlinewidthsasa functionofdistancein the
samewayasforourFig.4,i.e.,byperformingaGaussianfitto
each PV spectrum.We find that the broad wings of the Moffat
functionseverelysmearoutvelocitygradients.Spatialconvolu-
tion by the PSF is thus essential for a meaningful comparison
withobservedcentroids.Incontrast,spectralconvolutionaffects
onlythelinewidths.Forsimplicity,velocityvectorsareassumed
to have a constantmodulusV0 and to make a constantangleα Fig.8. Predicted emission maps for toy models (black contours) con-
from the local normal to the cavity. The flow velocity parallel volvedbyaGaussianof0(cid:3).(cid:3)12FWHM,superimposedontheobserved
andperpendiculartothecavitywallaregivenbyV(cid:13) ≡ V0sinα deconvolvedH2 1–0S(1)mapfromFig.1(redcontours).Top:hollow
andV⊥ ≡V0cosα,countedpositivelyawayfromthestar. cone viewed 40◦ from pole-on with radius 60 AU and semi-opening
ThepredictedPVcentroidvelocitiesinunitsofV areplot- angle27◦.Bottom:hollow half-sphere viewed45◦ frompole-on, with
0
tedinthetoppanelofFig.9asafunctionofdistancealongthe radius28AUandcenterproj√ectedat40AUalongthebluejet.Allcon-
toursincreasebyfactorsof 2.ThePSFofthespectrumat(0,0)was
jet, for both the cone and the half-sphere models presented in
Fig.8.Theresultsarestrikinglysimilardespitethedifferentge- subtractedinallcases.
ometries.Thisgivesusconfidencethatthe derivedconclusions
are robust and not dependent on the exact cavity shape. For a
pureparallelfloworientedawayfromthesource(cosα=0),the zeroobservedat–0(cid:3).(cid:3)4.Inthiscase,thedataarewellfittedwith
PVcentroidisblueshiftedatalldistancesby(cid:6) −0.5V(cid:13).Incon- V⊥ (cid:6) 20kms−1.AlargerV⊥ (cid:6) 30kms−1 wouldpredictexces-
trast,forapureperpendicularfloworientedoutward(cosα=1), siveblueshiftsclosetothesource.Examinationofthepredicted
the centroid is everywhere redshifted by an amount that in- linewidthsyieldsthesameresult,namelythatV⊥ (cid:6) 20kms−1
creases with distance up to (cid:6)+0.65V⊥. Centroids for interme- is compatible with the data while V⊥ (cid:6) 30 kms−1 predicts
diatevaluesofcosαaresimplygivenbyalinearcombinationof excessive line broadening.The maximum flow speed compati-
thesetwoextremes.Aparticularlyinterestingcaseiscosα=0.6 ble with the observed radial velocities and line widths is thus
(i.e.,V(cid:13)/V⊥ = 1.3),shownbythemiddlecurvesinFig.9.This V0 =V⊥/cosα(cid:6)30kms−1.
produces radial velocities that drop to essentially zero at –0(cid:3).(cid:3)4 If the –20km s−1 blueshiftat –0(cid:3).(cid:3)1 is instead due to an ax-
fromthestar,asobserved. ialjetknot,thevalueofαinthewide-angledrimsisnolonger
InthebottompanelofFig.9,theobservedH 1–0S(1)cen- well determined. The only constraint now is that PV centroids
2
troids in a PV along the jet (from Fig. 4) are compared with at–0(cid:3).(cid:3)2to–0(cid:3).(cid:3)4fallintherange–5to0kms−1 (withintheun-
thosepredictedforaconicalsurface(similarconclusionswould certainties discussed in Sect. 2). This approximately translates
be reached for the half-sphere). Only an intermediate value of into the condition (V(cid:13) −V⊥) (cid:6) 5 km s−1. For example, a pure
cosα (cid:6) 0.6(i.e.,V(cid:13)/V⊥ = 1.3)canreproduceatthesame time inward perpendicular flow with V⊥ (cid:6) −5 km s−1 (dash-dotted
the large blueshiftobservedat –0(cid:3).(cid:3)1, and the centroidsclose to curveinthebottompanelofFig.9)orapureparallelflowwith
A11,page10of17