Water in star-forming regions with Herschel

49 Pages·2014·3.34 MB·English

A&A572,A21(2014) Astronomy
DOI:10.1051/0004-6361/201424267 &
(cid:2)c ESO2014 Astrophysics
Water in star-forming regions with Herschel (WISH)
V. The physical conditions in low-mass protostellar outflows revealed
(cid:2),(cid:2)(cid:2),(cid:2)(cid:2)(cid:2)
by multi-transition water observations
J.C.Mottram1,L.E.Kristensen2,E.F.vanDishoeck1,3,S.Bruderer3,I.SanJosé-García1,A.Karska3,R.Visser4,
G.Santangelo5,6,A.O.Benz7,E.A.Bergin4,P.Caselli8,3,F.Herpin9,10,M.R.Hogerheijde1,D.Johnstone11,12,13,
T.A.vanKempen1,R.Liseau14,B.Nisini6,M.Tafalla15,F.F.S.vanderTak16,17,andF.Wyrowski18
1 LeidenObservatory,LeidenUniversity,POBox9513,2300RALeiden,TheNetherlands
e-mail:[email protected]
2 Harvard-SmithsonianCenterforAstrophysics,60GardenStreet,Cambridge,MA02138,USA
3 MaxPlanckInstitutfürExtraterrestrischePhysik,Giessenbachstrasse1,85748Garching,Germany
4 DepartmentofAstronomy,UniversityofMichigan,500ChurchStreet,AnnArbor,MI48109-1042,USA
5 OsservatorioAstrofisicodiArcetri,LargoEnricoFermi5,50125Florence,Italy
6 OsservatorioAstronomicodiRoma,viadiFrascati33,00040MonteporzioCatone,Italy
7 InstituteforAstronomy,ETHZurich,8093Zurich,Switzerland
8 SchoolofPhysicsandAstronomy,UniversityofLeeds,LeedsLS29JT,UK
9 UniversitédeBordeaux,ObservatoireAquitaindesSciencesdel’Univers,2ruedel’Observatoire,BP89,33270FloiracCedex,
France
10 CNRS,LAB,UMR5804,Laboratoired’AstrophysiquedeBordeaux,2ruedel’Observatoire,BP89,33270FloiracCedex,France
11 JointAstronomyCentre,660NorthAohokuPlace,UniversityPark,Hilo,HI96720,USA
12 DepartmentofPhysicsandAstronomy,UniversityofVictoria,POBox3055STNCSC,Victoria,BCV8W3P6,Canada
13 NRC-HerzbergInstituteofAstrophysics,5071WestSaanichRoad,Victoria,BCV9E2E7,Canada
14 DepartmentofEarthandSpaceSciences,ChalmersUniversityofTechnology,OnsalaSpaceObservatory,43992Onsala,Sweden
15 ObservatorioAstronómicoNacional(IGN),AlfonsoXII3,28014Madrid,Spain
16 SRONNetherlandsInstituteforSpaceResearch,POBox800,9700AVGroningen,TheNetherlands
17 KapteynAstronomicalInstitute,UniversityofGroningen,POBox800,9700AVGroningen,TheNetherlands
18 Max-Planck-InstitutfürRadioastronomie,AufdemHügel69,53121Bonn,Germany
Received24May2014/Accepted19September2014
ABSTRACT
Context.Outflowsareanimportantpartofthestarformationprocessasboththeresultofongoingactiveaccretionandoneofthe
mainsourcesofmechanicalfeedbackonsmallscales.Wateristheidealtraceroftheseeffectsbecauseitispresentinhighabundance
fortheconditionsexpectedinvariouspartsoftheprotostar,particularlytheoutflow.
Aims.Weconstrainandquantifythephysicalconditionsprobedbywaterintheoutflow-jetsystemforClass0andIsources.
Methods.Wepresentvelocity-resolvedHerschelHIFIspectraofmultiplewater-transitionsobservedtowards29nearbyClass0/Ipro-
tostarsaspartoftheWISHguaranteedtimekeyprogramme.ThelinesaredecomposedintodifferentGaussiancomponents,witheach
componentrelatedtooneofthreepartsoftheprotostellarsystem;quiescentenvelope,cavityshockandspotshocksinthejetandat
radex
thebaseoftheoutflow.Wethenusenon-LTE modelstoconstraintheexcitationconditionspresentinthetwooutflow-related
components.
Results.Wateremissionatthesourcepositionisopticallythickbuteffectivelythin,withlineratiosthatdonotvarywithvelocity,in
contrasttoCO.Thephysicalconditionsofthecavityandspotshocksaresimilar,withpost-shockH densitiesoforder105−108cm−3
2
andH Ocolumndensitiesoforder1016−1018 cm−2.H Oemissionoriginatesincompactemittingregions:forthespotshocksthese
2 2
correspondtopointsourceswithradiioforder10−200AU,whileforthecavityshocksthesecomefromathinlayeralongtheoutflow
cavitywallwiththicknessoforder1−30AU.
Conclusions.Wateremissionatthesourcepositiontracestwodistinctkinematiccomponentsintheoutflow;Jshocksatthebaseof
theoutflow or inthejet,and Cshocksinathinlayer inthecavitywall.Thesimilarityof thephysical conditions isincontrast to
off-sourcedeterminationswhichshowsimilardensitiesbutlowercolumndensitiesandlargerfillingfactors.Weproposethatthisis
duetothedifferencesinshockpropertiesandgeometrybetweenthesepositions.ClassIsourceshavesimilarexcitationconditionsto
Class0sources,butgenerallysmallerline-widthsandemittingregionsizes.Wesuggestthatitisthevelocityofthewinddrivingthe
outflow,ratherthanthedecreaseinenvelopedensityormass,thatisthecauseofthedecreaseinH OintensitybetweenClass0andI
2
sources.
Keywords.stars:formation–ISM:jetsandoutflows–ISM:molecules–stars:protostars
(cid:2) Herschel is an ESA space observatory with science instruments
(cid:2)(cid:2)(cid:2) ReducedspectraareonlyavailableattheCDSviaanonymousftp
providedbyEuropean-ledPrincipalInvestigatorconsortiaandwithim-
tocdsarc.u-strasbg.fr(130.79.128.5)orvia
portantparticipationfromNASA.
http://cdsarc.u-strasbg.fr/viz-bin/qcat?J/A+A/572/A21
(cid:2)(cid:2) Appendicesareavailableinelectronicformat
http://www.aanda.org
ArticlepublishedbyEDPSciences A21,page1of49
A&A572,A21(2014)
1. Introduction theprotostarevolves(Bontempsetal.1996),inparticularfrom
Class 0 to Class I, it is not clear if this is because the outflow
Molecular outflows are a ubiquitous and necessary part of the
actuallydecreasesinstrengthorsimplybecausethereislessen-
star formation process. They remove angular momentum and
velopematerialavailabletorevealitspresence.
material from the protostellar environment in a feedback pro-
cesswhichhelpstheprotostarformadiskandgainmassinthe Water is the ideal molecule to resolve these questions. It is
short-termwhileultimatelyconspiringwiththeinitialcorecon- theprimaryiceconstituentandoxygenreservoirinprotostellar
ditionstostarveitinthelongterm.Thusunderstandingoutflows envelopes, sublimates at dust temperatures above ∼100 K and
isat theheartof developinga true lawof star formationwhich can also be formed efficiently in the gas phase at temperatures
canpredictthestellaroutcomebasedoninitialcoreproperties. above a few 100 K (see van Dishoeck et al. 2013, and refer-
The classical tracer of such outflows, low-J CO (J ≤ 4), encestherein).Atshockvelocitiesabove∼10−20kms−1 itcan
traces material in a mixing layer which has undergone turbu- alsobesputteredfromthegrainmantles(seee.g.Jiménez-Serra
lententrainmentfromthequiescentenvelope(e.g.Canto&Raga etal.2008a;VanLooetal.2013;Neufeldetal.2014;Suutarinen
1991;Ragaetal.1995)withgastemperaturesoforder50−100K et al. 2014). It is therefore potentially present in relatively
(e.g.Yıldizetal.2013).Thiscarriesawayasignificantamount high abundancein the gas-phasein the cavity shock,wind and
ofmassfromtheenvelope,butatrelativelylowvelocitiesofor- shockswithinthejet.Wateralsohasalargedipolemomentand
der5−20kms−1andlikelyfrommaterialentrainedatsomedis- EinsteinAcoefficients,andthereforemoreintenselineemission
tancefromtheprotostar.Therefore,thepropertiesderivedfrom thanspecieswithsmallerdipolemoments.Evenforsubthermal
thisemissionmaynotaccuratelyreflectthetotalmomentum,an- excitation, where the number density is well below the critical
gularmomentumandkineticenergytransportofthesystem.In density, water lines can be more easily detected than emission
addition,itdoesnottracetheactivesurfacewheretheenvelope fromspeciessuch as CO. The favourablecombinationof these
is currently being sculpted (e.g. Nisini et al. 2010; Santangelo factors makes water a good tracer of the kinematics of these
etal.2013)andsodoesnotprobethetruefeedbackconditions. regions.
In contrast, protostellar jets, as traced in atomic gas or Theexpectedkinematicsignaturesarerelatedtotheproper-
shocked H (e.g. Reipurth et al. 2000), are more directly tiesof theshocksin the outflowandjet (see,e.g.Draine1980;
2
linked with accretion onto the central protostar (e.g. Pudritz Hollenbach 1997). In discontinuous, “jump” (J-type) shocks,
et al. 2007; Shang et al. 2007). This material is moving faster thereis a sharpincrease in the accelerationof gasin the shock
(100−1000kms−1,Franketal.2014)andisathighertempera- withrespecttotheambientun-shockedmaterialbythepassage
turesbutlowerH numberdensitiescomparedtotheoutflow(of of the shock front. The line-centre of the emission from these
2
order103−104Kandn ∼103−104cm−3respectively,Bacciotti moleculesisthereforeshiftedfromthesourcevelocitytosome
H
& Eislöffel1999),and thereforehas highermomentumand ki- fraction of the shock velocity dependenton the viewing angle.
netic energy but lower mass. So-called “bullets” can be seen Thedistributionofvelocitiesinthepost-shockmaterial,andthus
in molecular species such as CO, SiO and more recently H O theFWHMoftheemissionline,willalsobedifferentfromthat
2
(Bachilleretal.1990,1991;Hiranoetal.2006;Santiago-García oftheambientmaterial.Alternatively,in“continuous”(C-type)
etal.2009;Kristensenetal.2011),wherematerialiscompressed shocks,themoleculesaresmoothlyacceleratedbytheshockand
(and thus cools more efficiently) due to shocks within the jet. so emission extendsfromthe sourcevelocityto thevelocityof
However,the pencil-beamnatureof the jet meansthatit is un- the shock. Therefore,for the same shock geometrylarger line-
likely to be a major factor in the disruption of the envelope as widthsareexpectedforC-typeshocks.HybridC-J-typeshocks
the protostar evolves from Class 0 (T < 70 K) to Class I can be formed if the shock conditions are such that a C-type
bol
(70≤T <650K;Lada&Wilking1984;Andreetal.1993). shock does not have time to reach steady-state; in this case, a
bol
Between these two extremes are two intermediate regions; J-typefrontdevelopsatthetimetheshockistruncated(Chieze
theoutflowcavitywhichmaybefilledwithawindwhichhasa etal.1998).Forsimplicity,intheremainderofthepaperwewill
similarorlargerdensitythanthejet(Panoglouetal.2012),and onlyrefertoCandJshocksgiventhetimedependantnatureof
the active cavity shockat the boundarybetweenthe cavity and C-J shocks. Multiple discrete shocks with different conditions
the quiescentenvelope(e.g. Velusamyet al. 2007;Visser et al. ororientationswith respectto theline of sightwillgiverise to
2012). For the latter, the gas temperature and H number den- multipleemissionlinecomponents.Itisalsopossiblethatboth
2
sity are of order 300−1000 K and 105−107 cm−3 respectively, CandJtypeshocksexistaspartofthesamestructure(e.g.,see
as traced by H O and high-J CO (e.g. Goicoechea et al. 2012; Fig.9ofSuutarinenetal.2014),inwhichcasethephysicalcon-
Karskaetal.20214;Kristensenetal.2013),whilethedustisbe- ditionswillbesimilarbutthetwoshockswillproducedifferent
low 100 K. Studies based on Herschel PACS (Poglitsch et al. lineprofiles.
2010)and/orSPIRE(Griffinetal.2010)observationsseemulti- Strong,broadandcomplexlineprofileshavebeenobserved
pledistincttemperaturecomponentsinCOexcitationdiagrams: inwatertowardsClass0andIprotostars(Kristensenetal.2010,
cold emission (∼100 K) for J < 14, warm emission (∼300 K) 2012), most recently using the Herschel Space Observatory
for 14 (cid:2) J (cid:2) 24 and sometimes also hot emission (∼750 K) (Pilbratt et al. 2010) as part of the “Water in star-forming re-
for J (cid:3) 24with the columndensitydecreasingwith increasing gions with Herschel” (WISH; van Dishoeck et al. 2011) guar-
temperature(Manojetal.2013;Karskaetal.2014;Greenetal. anteedtimekeyprogramme.Thesehavebeencomplementedby
2013). However, these observations are spectrally unresolved, spectraatoff-sourcepositionsalongseveralpromenantoutflows
whichmakesrelatingthesetemperaturecomponentstophysical (Santangelo et al. 2012, 2013; Vasta et al. 2012; Nisini et al.
partsoftheprotostarmorechallenging. 2013). While Kristensen et al. (2012) looked at the dynamical
Therefore,howthephysicalandexcitationconditionsinthe componentsforonewaterline,the1 −1 ground-statetransi-
10 01
differentpartsofprotostellaroutflow-jetsystemsarerelated,and tion at 557GHz, this paper seeks to use multiple water transi-
how these vary both with distance from the central protostar tions to probe the excitation conditionsand water chemistry in
andbetweensources,isnotcurrentlywellunderstood.Inaddi- thesesources.Inaddition,studyinghowtheexcitationofwater
tion,whileoutflowline-widthandforcedecreaseonaverageas variesbetweenthedifferentphysicalcomponentsisrequiredto
A21,page2of49
J.C.Mottrametal.:Waterinstar-formingregionswithHerschel(WISH).V.
disentanglethetemperaturecomponentsseeninspectrallyunre- Para Ortho
solvedPACS/SPIREobservations. 300 322 H2O(GHz)
Thegoalsofthispaperaretherefore:tousethemultipletran- H128O(GHz) 312
asistipoanrstooffwthaeteWroIbSsHervsuerdvteoywtaordidselnotwif-ymtahsespChlyasssic0a/lIcpornodtoitsitoanrss K) 200 313 220 2211153 303 10190796
(
present where water is emitting within low-mass protostellar y 2
outflows,tounderstandthedifferencesandsimilaritiesbetween erg 100 202 11987852 212
tthoeexcWopenlodbrieetgihoinnowswinitthhtihsaecbhvraaiernifgodeuessswpcaritirhptstsiooonfutrohcfeetjheeveto-solauumttifloponlew.asnydstoebmse,ravnad- En 0 111 99500011111302 110555748 1011670JKaKc
tionsusedforthisstudyinSect.2.Next,wepresentourresults
3 2 1 0 0 1 2 3
in Sect. 3 and additional analysis in Sect. 4. We then discuss K K
c c
theimplicationsofthese resultsin Sect.5,bothin termsofthe
different parts of the jet-outflow system (Sect. 5.1), the impact Fig.1. Level diagram of the various H O (red) and H18O (blue) tran-
2 2
of source evolution (Sect. 5.2), and comparison to off-source sitions observed with HIFI towards the WISH sample of low-mass
shocks(Sect.5.4).Finally,wesummariseourmainfindingsand protostars.
reachourconclusionsinSect.6.
baselinewasthenusedtocalculatethecontinuumlevel,compen-
2. Observations
satingforthedual-sidebandnatureoftheHIFIdetectorsi.e.the
The WISH low-mass sample consists of 15 Class 0 and initialcontinuumlevelisthecombinationofemissionfromboth
14ClassIsources,thepropertiesofwhicharegiveninTable1. the upper and lower sideband, which we assumed to be equal.
Allsourceshavebeenindependentlyverifiedastrulyembedded FollowingthistheWBSsub-bandswerestitchedintoacontinu-
sourcesandnotedge-ondisks. ousspectrumandalldatawereconvertedtotheT scaleusing
MB
Thissamplewasthetargetofaseriesofobservationsofgas- efficiencies from Roelfsema et al. (2012). Finally, for ease of
phase water transitionswith the HeterodyneInstrumentfor the analysisall data were convertedto FITSformatand resampled
Far-Infrared(HIFI;deGraauwetal.2010)onHerschelbetween to0.3kms−1spectralresolutiononthesamevelocitygridusing
March 2010 and October 2011. Three of the Class I sources bespokepythonroutines.
(IRAS3A, RCrA-IRS5A and HH100-IRS) were only observed
Comparisonofthetwopolarisationsforeachsourcerevealed
in the 557GHz H2O 110−101 line, which was presented for insignificant differences, so these were co-added to reduce the
all sources by Kristensen et al. (2012). All other sources were
noise. Comparison of peak and integrated intensities between
observed in between four and seven H16O transitions and be-
2 the original WISH observations and those obtained as part of
tweenoneandfourH18Otransitions.Additionaldatafromtwo OT2_rvisser_2forthesamesourcessuggestthatthecalibration
2
OT2programmes,OT2_rvisser_2andOT2_evandish_4,arealso uncertainty is (cid:2)10%. For the 2 −1 line for BHR71, the off-
02 11
includedtoaugmenttheWISHdata. positions of the DBS mode coincided with outflow emission,
Detailsofthelinefrequency,main-beamefficiency,spectral resulting in a broad absorption. This is masked out during the
andspatialresolutions,observingtime,criticaldensityat300K analysis so does not impact the results for this source. In ad-
andupperlevelenergyoftheobservedtransitionsaregivenfor dition, as also noted for the 1 −1 transition by Kristensen
10 01
alllines in Table A.1.SettingsprimarilytargetingH18O transi- et al. (2012), observations of the three Serpens sources some-
2
tionswereonlyobservedtowardsClass0sourceswherehigher timesshowaweaknarrowabsorptionfeatureatv =1kms−1
LSR
lineintensitieswereexpectedcomparedtoClassIsources.This whichprobablyarrisesfromemissioninthereferenceposition.
alsomotivatedthelongerintegrationsintheH2O 111−000 tran- Thisdoesnothaveanyimpactontheresultsderivedbelowand
sition for Class 0 than Class I sources as that setting also in- soisignored.
cludes the corresponding H18O transition. Longer integrations InfivesourcestheC18OJ =10−9lineisdetectedintheline
2
were performedforClass I sourcesin the 110−101 transition to wing of the H2O 312−303 (1097GHz) line. Before performing
ensuredetectionsinatleastonelineinthemaximumnumberof analysis on these data, we remove the C18O emission by sub-
sources.AleveldiagramofthevariouslinesisshowninFig.1 tractinga Gaussianwith thesame FWHM, line-centreandam-
and the observations identification numbers of all data used in plitudeasobtainedbySanJosé-Garcíaetal.(2013).
thispaperaregiveninTableA.2. As noted in Table 1, the more accurate SMA coordinates
All observationswere taken in both horizontaland vertical for IRAS 15398 were observed in two settings, H O 1 −0
2 11 00
polarisationswithboththeWideBandSpectrometer(WBS)and and H18O 1 −1 , as part of programme OT2_evandish_4.
High Resolution Spectrometer (HRS) backends. Observations Compa2rison1o0f th0e1se observationswith the WISH observations
weretakenassinglepointingsindual-beam-switch(DBS)mode is discussed in Appendix B.1. In the rest of this paper we will
with a chop throw of 3(cid:5), with the exception of some of the focusontheWISHobservationsastheseincludethemosttran-
H2O110−101 observations,whichweretakeninposition-switch sitionsobservedtowardsthesameposition.
mode(seeKristensenetal.2012,formoredetails).TheHerschel
beamrangesfrom12.7(cid:5)(cid:5)to38.7(cid:5)(cid:5)overthefrequencyrangeofthe
variouswaterlines, closeto thediffractionlimitoftheprimary
3. Results
mirror.
hipe
The data were reduced with (Ott 2010). After initial Thissectionbeginswithpresentationofthoseresultsthatcanbe
spectrumformation,furtherprocessingwas also performedus- obtainedsimplyfrom the data themselves(Sect. 3.1).The pro-
hipe
ing . This began with removal of instrumental standing files are then fitted with multiple Gaussian components,which
waves where required, followed by baseline subtraction with aresubsequentlydividedintodifferentphysicallymotivatedcat-
a low-order (≤2) polynomial in each sub-band. The fit to the egoriesbasedontheirproperties(Sect.3.2).
A21,page3of49
A&A572,A21(2014)
Table1.Sourceparameters.
Source RA Dec Da (cid:2) b L c T c M d F e
LSR bol bol env CO
(hms) (◦(cid:5)(cid:5)(cid:5)) (pc) (kms−1) (L(cid:7)) (K) (M(cid:7)) (M(cid:7)yr−1kms−1)
L1448-MM 032538.9 +304405.4 235 +5.2 9.0 46 3.9 3.7×10−3
NGC1333-IRAS2A 032855.6 +311437.1 235 +7.7 35.7 50 5.1 7.4×10−3
NGC1333-IRAS4A 032910.5 +311330.9 235 +7.2 9.1 33 5.2 2.1×10−3
NGC1333-IRAS4B 032912.0 +311308.1 235 +7.4 4.4 28 3.0 2.2×10−4
L1527 043953.9 +260309.8 140 +5.9 1.9 44 0.9 4.4×10−4
Ced110-IRS4 110647.0 −772232.4 125 +4.2 0.8 56 0.2 −
BHR71 120136.3 −650853.0 200 −4.4 14.8 44 3.1 −
IRAS15398f 154301.3 −340915.0 130 +5.1 1.6 52 0.5 9.5×10−5
L483 181729.9 −043939.5 200 +5.2 10.2 49 4.4 5.9×10−4
Ser-SMM1 182949.8 +011520.5 415 +8.5 99.0 39 52.5 3.0×10−3
Ser-SMM3 182959.2 +011400.3 415 +7.6 16.6 38 10.4 4.2×10−3
Ser-SMM4 182956.6 +011315.1 415 +8.0 6.2 26 6.9 4.8×10−3
L723 191753.7 +191220.0 300 +11.2 3.6 39 1.3 2.9×10−3
B335 193700.9 +073409.6 250 +8.4 3.3 36 1.2 6.0×10−4
L1157 203906.3 +680215.8 325 +2.6 4.7 46 1.5 3.7×10−3
NGC1333-IRAS3A 032903.8 +311604.0 235 +8.5 41.8 149 8.6 −
L1489 040443.0 +261857.0 140 +7.2 3.8 200 0.2 1.6×10−4
L1551-IRS5 043134.1 +180805.0 140 +6.2 22.1 94 2.3 5.1×10−4
TMR1f 043913.7 +255321.0 140 +6.3 3.8 133 0.2 2.5×10−5
TMC1Af 043934.9 +254145.0 140 +6.6 2.7 118 0.3 1.3×10−4
TMC1 044112.4 +254636.0 140 +5.2 0.9 101 0.2 4.5×10−4
HH46-IRS 082543.9 −510036.0 450 +5.2 27.9 104 4.4 1.1×10−3
IRAS12496 125317.2 −770710.6 178 +3.1 35.4 569 0.8 −
GSS30-IRS1 162621.4 −242304.0 125 +3.5 13.9 142 0.6 5.2×10−4
Elias29 162709.4 −243719.6 125 +4.3 14.1 299 0.3 6.4×10−5
Oph-IRS63 163135.6 −240129.6 125 +2.8 1.0 327 0.3 1.1×10−5
RNO91 163429.3 −154701.4 125 +0.5 2.6 340 0.5 1.0×10−4
RCrA-IRS5A 190148.0 −365721.6 130 +5.7 7.1 126 2.0 −
HH100-IRS 190149.1 −365816.0 130 +5.6 17.7 256 8.1 −
Notes.SourcesabovethehorizontallineareClass0,sourcesbelowareClassI.(a) TakenfromvanDishoecketal.(2011)withtheexceptionof
sourcesinSerpens,whereweusethedistancedeterminedusingVLBAobservationsbyDzibetal.(2010).(b)Obtainedfromground-basedC18O
orC17Oobservations(Yıldizetal.2013)withtheexceptionofIRAS4AforwhichthevaluefromKristensenetal.(2012)ismoreconsistentwith
ourdata.(c)MeasuredusingHerschelPACSdatafromtheWISHandDIGITkeyprogrammes(Karskaetal.2014).(d)Masswithinthe10Kradius,
determinedbyKristensenetal.(2012)fromdustymodellingofthesources.(e)TakenfromYıldizetal.(2014)forCO3−2.(f)Thecoordinates
usedinWISH;moreaccurateSMAcoordinatesofthesourcesare15h43m02s.2,−34◦09(cid:5)06(cid:5).(cid:5)8(IRAS15398),04h39m13s.9,+25◦53(cid:5)20(cid:5).(cid:5)6(TMR1)
and04h39m35s.2,+25◦41(cid:5)44(cid:5).(cid:5)4(TMC1A;Jørgensenetal.2009).ForIRAS15398,thesecoordinateswereobservedintwosettingsaspartofthe
OT2programmeOT2_evandish_4.
3.1.Lineprofiles from the source velocity that are above 2σ of the resam-
rms
pled spectrumwithin a window aroundthe line are found.The
All H O spectra for three Class 0 and two Class I sources
2 FWZI is then between the first channelmoving away from the
are shown in Fig. 2 as an example, with spectra for all WISH
sourcevelocityin each directionwhere thespectrumdropsbe-
sourcespresentedinFigs.A.1−A.6forallH Otransitions.The
2 low 1σ . The integrated intensity is then calculated over the
H18O spectra for sourceswith at least one detectionare shown rms
2 rangeidentifiedbytheFWZI.Whilethisapproachismoredata
inFig.A.7.
than source driven, there is approximately a factor of 10 dif-
As can be seen in Fig. 2, the water line profiles are often
ference in the noise level between the deepest and shallowest
broad and complex, with generally narrower emission towards
spectra (see Table 2). Thus using an alternative definition of
ClassIwithrespecttoClass0sources.Thereissignificantvari-
the FWZI based on a set fraction of the peak in a way that
ationinlineintensityandshapebetweendifferentsources,which is consistent and comparable between the different transitions
is notparticularlysurprisinggiventhe rangethe sample covers
wouldrequirea highenoughthresholdthatitwouldnotreflect
intermsofluminosity,envelopemassandoutflowactivity(see
the broadnessof the line wings. It could also be skewed in the
furtherdiscussioninSect.4.3). lowerexcitationlinesbythenarrowemissionand/orabsorption
Thebasicpropertiesofthespectra;noiselevelin0.3kms−1
atthesourcevelocity(forexample,seeIRAS4BinFig.2).
bins,peakbrightnesstemperature,integratedintensityandfull-
widthatzerointensity(FWZI),aretabulatedforallsourcesand Table2presentsthedetectionstatistics, mediannoiselevel,
linesinTablesA.3−A.5. andthemeanandmedianFWZI foralldetectionsseparatedby
TheFWZIismeasuredonspectraresampledto3kms−1 to the evolutionary stage of the source. For the Class 0 sources,
improvethesignal-to-noiseratio(S/N).First,thefurthestpoints BHR71 and L1448-MMare excludedbecause they have bullet
A21,page4of49
J.C.Mottrametal.:Waterinstar-formingregionswithHerschel(WISH).V.
IRAS4B Ser− SMM3 BHR71 Elias29 IRAS12496
1.0
×10 ×4
0.5
3 −3
12 03
0.0
1.0
×0.8 ×4 ×4 ×3
0.5
3 −2
12 21
0.0
1.0
×5 ×5 ×5 ×15
0.5
2 −2
11 02
0.0
1.0
×0.6 ×2
0.5
2 −1
12 01
0.0
1.0
×0.6 ×3 ×5 ×2.5 ×7
2 −1 0.5
02 11
0.0
1.0
×6 ×2 ×5 ×10
1 −1
10 01 0.5
0.0
1.0
111−000 ×0.7 ×5 ×5 ×3 ×5
)
K 0.5
(
B
M
T 0.0
−75 0 75 −75 0 75 −75 0 75 −75 0 75 −75 0 75
v(km s−1)
Fig.2.ExampleH OspectraforthreeClass0andtwoClassIsources(namesinredandbluerespectively).Allspectrahavebeenrecentredso
2
thatthesourcevelocityisat0kms−1andscaledbythenumberinthetop-rightcornerofeachpanel.Somespectrahavealsobeenresampledtoa
lowervelocityresolutionforeaseofcomparison.Thegreenlineindicatesthebaseline.
emission (discussed further in Sect. 3.2.4) which significantly suggests that on average our observations have a high-enough
increases their FWZI compared to other sources but were not sensitivity to detect the fullextentof the line wings. While the
observedinalllines. Class I sourcesare fainter and so have a lower S/N, the transi-
TheaverageH OFWZIs(seeTable2)areremarkablysim- tions also look narrower, so it seems unlikely that higher sen-
2
ilar for Class 0 sources. There is also little difference between sitivity would increase their mean FWZI to the point where it
the mean and median values, suggesting that these values are wasconsistentwiththeClass0sources.Variationinlineshape
notdominatedbyafewsourcesandsoarerepresentativeofgen- betweentransitionsforagivensourceisrelativelysmall,partic-
eralsourceproperties.Giventheorderofmagnitudedifference ularlyinthelinewingsfortheClass0sources.Inafewcasesthe
between the highest and lowest sensitivity observations, this FWZIvariesbetweenthedifferenttransitionsforagivensource,
A21,page5of49
A&A572,A21(2014)
Table2.Detectionstatistics,averagenoiseandFWZIforeachlineandevolutionarystage.
Line Class0 ClassI
D/O.a σ MeanFWZI MedianFWZI D/Oa σ MeanFWZI MedianFWZI
rms rms
(mK) (kms−1) (kms−1) (mK) (kms−1) (kms−1)
H O1 −0 14/15 19 79±32b 82b 7/11 24 47±18 42
2 11 00
H O1 −1 15/15 12 72±32b 69b 12/14 10 50±19 48
2 10 01
H O2 −1 5/5 123 69±13 63 0/0 − − −
2 12 01
H O2 −1 14/15 22 75±33b 81b 9/11 22 34±13 34
2 02 11
H O2 −2 12/15 20 65±27b 62b 7/9 17 33±15 33
2 11 02
H O3 −2 7/15 105 58±21b 54b 4/11 122 22±5 22
2 12 21
H O3 −3 8/8 17 81±26c 75c 2/2 9 42d 42d
2 12 03
H18O1 −0 1/15 18 16d 16d 0/11 26 − −
2 11 00
H18O1 −1 3/13 4 41±8 45 0/1 4 − −
2 10 01
H18O2 −1 0/3 16 − − 0/0 − − −
2 02 11
H18O3 −3 2/8 14 12d 12d 1/2 8 33d 33d
2 12 03
Notes.(a)No.ofsourceswithdetectionsoutofthetotalobservedineachline.(b)DetectionsforBHR71andL1448-MMexcluded.(c) Detection
forL1448-MMexcluded,BHR71notobserved.(d)Nostandarddeviationisgivenfordetectionsinlessthanthreesources.
but in all cases except Ser-SMM3 the results are due to varia-
tion in the noise level of the different spectra. The reason that H2O 110−101/24
Ser-SMM3 is likely still consistent with the general picture is 0.04 H18O 1 −1
discussedinAppendixB.2. 2 10 01
The FWZI for all H18O detections except the 3 −3 line
2 12 03
towards Elias29 are smaller than those for the corresponding
H16Otransitionbyafactorof2−8.However,asshowninFig.3,
th2e spectra are consistent within the noise. Thus the difference ) CH
K
tisiemsaossstulmikienlygaanS/iNsotiosspuice.1C6Oom/18pOarirsaotinooofft5h4e0in(tWegirlsaotend&intReonosid- (B 0.02 CH
M
1994)resultsinanopticaldepthfortheH Otransitionsoforder
2 T
20−30assumingthatH18Oisopticallythin.
2
Only TMC1A and Oph-IRS63, both Class I sources, were
not detected in any transition at the 3σ level (in 0.3 km s−1
bins).Allsourcesdetectedinthe1 −1 linearealsodetected, 0.00
10 01
whereobserved,inallotherH Olinesexceptthe3 −2 tran-
2 12 21
sition.Thenon-detectionsinthislinearelikelyduetothehigher
noise in these data as it is generally sources that are fainter in
the other linesthat are notdetected.Thisis also likely the rea- −40 −20 0 20 40
sonforthenon-detectionsinH128Oasitisonlytheverybrightest v(km s−1)
sourcesthataredetected,andeventhenmosthaveapeaksignal-
to-noiseoflessthan10.Ofthe14sourcesobservedintheH18O
2 Fig.3.ComparisonoftheH O(black)andH18O(red)1 −1 spectra
1 −1 transition,seven(BHR71,L1527,NGC1333-IRAS2A, 2 2 10 01
10 01 forIRAS4BwheretheH Ospectrumhasbeenscaleddownsuchthat
NGC1333-IRAS4A,NGC1333-IRAS4B,Ser-SMM1andSer- 2
thepeakintensitiesarethesame.Thegreenlineindicatesthebaseline.
SMM4)havedetectionsoftheCHtripletat536.76−536.80GHz ThebluelinesindicatetheapproximatevelocitiesoftheCHtransitions
inemissionintheotherside-band.ForNGC1333-IRAS4A,this fromtheothersidebandintheH18Oobservations.Thethirdcomponent
isconfusedwiththeH18Oline,sotheCHtripletismaskeddur- oftheCHtripletisjustbeyondt2heplottedrangebutisalsodetected.
2
ing the analysis. Analysis of the CH emission itself is beyond
thescopeofthispaper. profile. However, as shown by Kristensen et al. (2010, 2012)
The conclusion from the comparison of line profiles and theycanbedecomposedintomultiplecomponents,eachrelating
FWZIsisthereforethatthelowerFWZIfortheH18Otransitions todifferentpartsoftheprotostellarsystem.
2
comparedto the correspondingH16O line for a givensource is Inreality,thedetailedshapeoftheemissionfromagivenre-
2
justaS/Nissue.However,thedecreaseintheaverageFWZIbe- gionwilldependonboththephysicsandgeometry,particularly
tween Class 0 and I is real and not related to the sensitivity of forshocks,andsoarangeoflineshapesmayindeedbepresent
thedata. (see e.g. Jiménez-Serra et al. 2008b). However, the observed
H Oline-shapes,particularlyinhighS/Ndata,appearGaussian-
2
like, so this is the most reasonable line-shape to assume. The
3.2.Linecomponents reason that the emission from shocks is Gaussian-like may be
duetoourobservationsencompassinganumberofshockswith
3.2.1. Gaussiandecomposition
a rangeof viewingangles.Alternatively,this may be the result
AscanbeseeninFig.2,thewaterlineprofilestowardslow-mass ofmixingandturbulenceinducedbyKelvin-Helmholtzinstabil-
protostarsare complexand generally notwell reproducedby a itiesalongthecavitywall(seee.g.Bodoetal.1994;Shadmehri
singlelineshape,e.g.asingleGaussian,Lorentzianortriangular &Downes2008).
A21,page6of49
J.C.Mottrametal.:Waterinstar-formingregionswithHerschel(WISH).V.
As discussed in Sect. 3.1, the width of the line profiles 0.4
doesnotchangesignificantlybetweenthe observedtransitions, 1 −1 1 −0
10 01 11 00
thoughthe relative andabsolute intensityof individualcompo- ×2
nents does change. Therefore, while the physicalconditions in 0.2 S S
thedifferentregionswithintheprotostarwherewaterisemitting
may be different,all transitionsare probablyemitting from the
sameparcelsofgasineachcase. 0.0
Wethereforechoosetorequirethatthelinecentreandwidth
of each Gaussian componentare exactly the same for all tran-
sitionsobservedtowardsagivensource,thoughtheintensityof 2 −1
02 11
a given component can be different for each line. In practical ×3 C
terms, this is achieved by creating an array which contains all
H O and H18O spectra for a given source and fitting a global
2 2
function to this array which contains a number of Gaussians
equaltothenumberofcomponentsmultipliedbythenumberof
transitions.Foragivencomponent,thelinecentreandwidthare
commonvariablesbetweentheGaussiansappliedto each tran- 0.4
sition. They are therefore constrained by all available data for 3 −2 2 −2
12 21 11 02
a particular source, decreasingthe uncertaintiesand improving ×1.5 ×3 S
thereliabilityofthefit,particularlyincaseswheretheemission K) 0.2
insometransitionsisweak.ForhighS/Nspectra,thedifference
(
between fitting each line separately and this global fitting ap- B
M
proachis small, as shown in Fig. 2 of Kristensen et al. (2013). T
0.0
Those authorswere able to use individualfits because they fo-
cusedonthebrightestClass0sourcesintheWISHsampleand
were interested in one relatively distinct component. Here we −70 0 70 −70 0 70
want to isolate and analyse all componentsin all sources, so a v(km s−1)
globalfittingapproachispreferred.
An examplebest-fit resultis shownin Fig. 4 for BHR71, a
sourcewithamixoflowandhighS/Nspectra.Forthissource, Fpliegd.4to.C3oknmtinsu−u1m.TshuebrteradcatenddWcyBanSlsipneecstsrahofwortBheHiRnd7i1vi(dbulaaclkG)aruesssaiman-
the quiescent envelope component shows an inverse P-Cygni
components for thecavityshock (C)andspot shocks (S)respectively
profile at full resolution and so is masked out from the fitting
(seetextandTable3fordetails)whilethebluedashedlineshowsthe
process.Inothercaseswhereonlyasimpleemissionorabsorp- combined fit for each line. All spectra have been shifted so that the
tionprofilefromtheenvelopeisobserved,thisisincludedinthe sourcevelocityisat0kms−1,whichisindicatedbythegreendashed
Gaussianfit. Forthe2 −1 transitionthe absorptionisdueto lines. At full resolution the quiescent envelope component has an in-
02 11
referencecontaminationandsoisalsomaskedfromthefitting. verseP-cygni profile(see Mottramet al.2013) and soismasked (in-
Thefitresultswereobtainedusingtheordinaryleast-squares dicatedbythemagentadashedlines)ratherthanbeingfitbymultiple
solverin thepythonmodulescipy.odr1 startingfroman ini- components during the Gaussian fitting. The broad absorption in the
2 −1 transition(middleleftpanel)iscausedbyreferencecontamina-
tialguessforasingleGaussian.Theresultsandresidualsofthis 02 11
tionandisalsomaskedduringthefittingprocess.
fit were examined and the number of componentsincreased or
the initial guess modified to result in residuals below the rms.
Whilethisapproachcanbesusceptibletofindinglocalminima Table3.Componentterminology.
insomecases,particularlywithverycomplexlineprofilessuch
asforBHR71,thecombinationofvaryingtheinitialguessand
Thispapera Previouspapers References
visualinspectionoftheresidualsensuredthatthisreturnedrea-
Envelope Narrow 1
sonableresults(e.g.combinationsoflargepositiveandnegative
Cavityshock Broadormedium 1
Gaussians which mostly cancel out are excluded). In all cases
Spotshock BulletorEHV,alsooffset 1,2
thenumberofGaussiancomponentsusedwastheminimumre-
ormediumifbroadalsopresent
quiredfortheresidualstobewithinthermsnoise.Theresultsof
theGaussianfittingforallsourcesarepresentedinTablesA.6to Notes.(a)SeeSects.3.2.2−3.2.4andFig.6forcriteria.
A.10.Where a componentis notdetected in a givenline, a 3σ
References.(1)Kristensenetal.(2012);(2)Kristensenetal.(2013).
upper limit is calculated from the noise in the spectrum. The
results are consistent with those presented in previous papers
(Kristensen et al. 2012,2013;Mottram et al. 2013)taking into
accountthelatestreductionandcalibration. distinctionbetweendifferentexcitationconditions,asalsonoted
Having identified these components, it is then a question in Kristensen et al. (2013). We therefore prefer to use terms
of attempting to relate them to the different physical compo- which indicate the most likely physical origin of the emission
nentsofaprotostellarsystem.Inpreviouswork(Kristensenetal. component(cf.vanderTaketal.2013,forsimilarterminology
2010,2012;SanJosé-Garcíaetal.2013;Yıldizetal.2013)the appliedtohigh-massprotostars).Table3providesasummaryof
different components have been established and named based howthesenewtermsarerelatedtothoseusedinpreviouspapers
primarily on their line-width. However, this is a rather phe- onlow-massprotostarsinordertoensurecontinuity.
nomenologicalconventionand does not always allow for clear The different components for each source are divided into
three categories: envelope,cavity shock and spot shock, build-
1 http://scipy.org/ ingontheworkofKristensenetal.(2012,2013),withthefirst
A21,page7of49
A&A572,A21(2014)
parts of the spectra are masked during the fitting process (e.g.
seeFig.4).
Absorption from the envelope is also observed in the
H18O 1 −0 and1 −1 linestowardsSMM1,whichiscon-
2 11 00 10 01
sistent with the envelopeof this sourcebeingparticularlymas-
siveandhavingarelativelyshallowdensitypower-lawslope(cf.
Kristensenetal.2012).Theonlysourcetoshowemissionfrom
theenvelopeinanyH18OtransitionsisIRAS2A,wheretheten-
2
tative detectionsin the H18O 3 −3 and2 −1 lines are the
2 12 03 02 11
narrowestforanysourceandoffsetfromthemainoutflowemis-
siondetectedintheH16Otransitions.Thisemissionislikelyre-
2
latedtothehotcorewhereT >100K(seeVisseretal.2013,
dust
for more details) and originates on arcsecond scales based on
interferometricobservations(Perssonetal.2012,2014).Wedo
not study the envelope emission further in this paper. Further
analysisofothersourcesshowingabsorptionintheground-state
waterlines,includingthelinkwithwaterice,willbepresented
inSchmalzletal.(2014).
Fig.5.Cartoonshowingtheproposedoriginofthevariousdistinctkine- 3.2.3. Cavityshock
maticgascomponentsobservedinlow-Jwaterlineprofiles.
Having identified any envelope contribution, we designate the
remaining component which is not in absorption in any line
and has the smallest ratio of offset to FWHM as the cavity
shock component.This is an empiricaldeterminationbased on
letter of each term being used to identify them in Tables A.6
toA.10.Thefollowingsubsections(Sect.3.2.2−3.2.4)willdis- the assumption that the average velocity offset of the currently
shocked gas in the outflow cavity is lower than for more dis-
cussandmotivatethedefinitionofeachofthesecomponentsin
crete and energetic shocks and that it should not be in absorp-
turn,withFig.5indicatingtheirexpectedphysicallocationina
tion againstthe continuumbecause the emission is most likely
protostellarsystem.Followingthis,asummaryandcomparison
showing how the kinematic properties of the different compo- formedonlargerscales.Theoffsetofthiscomponentisalways
lessthan15kms−1anddecreaseswithsmallerFWHM.Thatthis
nentsrelatetoeachotherwillbepresentedtoverifythattheyare
componentisGaussianinshape,combinedwiththesmalloffset
distinct(Sect.3.2.5).
comparedtotheFWHM,suggeststhatwearedetectingboththe
redandblue-shiftedlobesoftheoutflowcavity.
3.2.2. Envelope
Water emission is elongatedalong the direction of the out-
flow (e.g. Nisini et al. 2010; Santangelo et al. 2012) with the
Emissionfromthequiescentenvelopeischaracterisedbysmall
FWHMandoffsetfromthesourcevelocity,thusweassignthis dominantextendedcomponenthavingsimilarvelocitydistribu-
tions(e.g.Santangeloetal.2014)asthiscomponent.Ascavity
designationtothecomponentwiththesmallestFWHMforeach
sourcewhichhasFWHM≤5kms−1andoffset≤2kms−1.This shocksalsodominatetheon-sourcelineprofiles,weassumethat
it is elongated along the outflow direction but does not fill the
can be in absorption in the ground-state lines, particularly for
beamparalleltotheoutflowaxis,asinthespectrallyunresolved
Class 0 sources, and even saturated where all line and contin-
PACSH Oobservations.
uumphotonsareabsorbed.Oneconfirmationthatthisemission 2
andabsorptioncomesfromtheenvelopeisthatthelinecentres This componentshould not be confused with the entrained
andwidthsaresimilartothoseobservedinC18Otowardsthese outflowmaterialtypicallyprobedbylow-J COobservations,as
sources (San José-García et al. 2013). No sources show dis- H O and low-J CO emission are not spatially coincident (e.g.
2
tinctforegroundabsorptionsoffsetfromthesourcevelocity,un- Nisini et al. 2010;Santangelo et al. 2013). A detailed compar-
likeHIFIspectratowardshigh-massprotostars(e.g.vanderTak ison between CO J = 3−2 and H O 1 −1 was presented in
2 10 01
etal.2013),primarilyduetothemuchsmallerdistancestoour Kristensenetal.(2012)withtheclearconclusionthatthesetwo
sources.Thusmostoftheabsorptionlikelycomesfromthepro- transitionsdonottracethesamematerial.Oneofthemainrea-
tostarsownenvelope.Giventhatthesub-mmcontinuumandline sonsistheenormousdifferenceincriticaldensitybetweenthese
emission from the envelope is centrally condensed (Jørgensen twotransitions(104cm−3and107cm−3respectively).Asthegas
etal.2007;Kristensenetal.2012;Mottrametal.2013)weas- is heatedandcompressedin the cavityshocks,the water abun-
sumethattheemissionscalesasapoint-source. dance increases dramatically through both gas-phase synthesis
While many sources also show envelope emission, it is andice sputtering.Duringthiswarm anddensephase,water is
often non-Gaussian in shape in the ground-state lines, con- oneofthedominantcoolants.However,asthegascoolsandex-
sisting of combinations of emission and absorption in either pandsto come into pressure equilibriumwith its surroundings,
inverse or regular P-Cygni profiles which are indicative of water excitation becomeshighly inefficientdue to high critical
infallandexpansionrespectively.Thiswascharacterisedin the densities so little water emission originates from the cold en-
1 −1 (557GHz)linebyKristensenetal.(2012),andthecases trained low-density outflow. Therefore, the non-coincidenceof
10 01
showinginfallprofileswereanalysedinmoredetailbyMottram waterandlow-J CO isconsistentwiththeexpectationthatwa-
etal.(2013).Inthesecases,thecombinationofenvelopeemis- ter is significantly depleted under the typical conditions in the
sionandabsorptionisnotasingleGaussianandsotherelevant entrainedoutflowinggas.
A21,page8of49
J.C.Mottrametal.:Waterinstar-formingregionswithHerschel(WISH).V.
MostdetectionsintheH18Oobservationsareassociatedwith
2 Gaussian
the cavity shock component,with the exceptionof IRAS2A as component
discussedaboveandIRAS4A,whichisdiscussedinmoredetail
inAppendixB.3.
3.2.4. Spotshock Lowest FWHM
Allremainingcomponentswhichshowlargeroffset/FWHMare Yes
designatedasspotshockcomponents.Theseparationofthecav- No FWHM ≤ 5 kms-1 Yes
ityandspotshockcomponentsisnecessarybecausethelinepro- |vpeak-vLSR| ≤ 2 kms-1 Envelope
filesshowseparateanddistinctkinematiccomponents(e.g.see No
Fig.4),suggestingthattheycomefromdifferentshockswithin
Lowest
the protostellar system. The use of offset/FWHM is also cho- |vpeak-vLSR| / FWHM, Yes
Not in absorption
sensoastoseparatethecomponentmostlikelyassociatedwith
C-type shocks(cavity shock), where emission is centred at the Cavity Shock
sourcevelocity,withcomponentsmorelikelyassociatedwithJ- No
type shocks(spotshock),where emission is shifted away from
the source velocity to the shock velocity relative to the line of Spot Shock
sight(seee.g.Hollenbach1997).
Some spot shock components are significantly offset from Fig.6.Flowdiagramforcomponenttypedetermination.|vpeak−vLSR|is
theoffsetofthecomponentcentrefromthesourcevelocity.
the sourcevelocity,suchthat theyare characteristicof “bullet”
emissionwithlargeoffsets(>20kms−1)fromthesourceveloc-
ityandlargeFWHM(also>20kms−1,e.g.seeFig.4).Theseare
mostlikelyassociatedwith J-typeshocksalongthejet,asthey
havesimilarkinematicpropertiestoEHVbulletemissioninCO
point-like knots in interferometric observations (e.g. Hirano
andSiOwhichisspatiallylocatedinknotsalongthejetaxis(e.g.
et al. 2006; Santiago-García et al. 2009) and the analysis of
Bachilleretal.1990,1991;Hiranoetal.2006;Santiago-García
Kristensenetal.(2013)suggeststhatthenon-bulletspotshocks
etal.2009). originate from very small regions (∼100 AU) near the central
Thespotshockemissionwithlowervelocityoffsetmayorig-
protostar.Thisisalsosupportedbythestrongsimilarityin line
inate in J-type shocks near the base of the outflow where the
shapebetweenthespotshockcomponentobservedinwaterfor
wind first impacts the envelope or outflow cavity, as first sug- IRAS2A and the compact (∼1(cid:5)(cid:5)) emission seen in SiO and SO
gestedbyKristensenetal.(2013).Thoseauthorsbasedthiscon-
towardsMM3inrecentinterferometryobservationsbyCodella
clusion on: (i) some of the spot shock componentsdetected in
etal.(2014).Apoint-likegeometryisthereforethemostappro-
waterlineprofilesareseeninabsorptionagainstthecontinuum
priateassumptionforthespotshockcomponent.
but not the outflow; (ii) when detected in OH+ and CH+, the
componentsarealwaysinabsorptionagainstthecontinuumwith
no emission componentand no outflow component.These two
3.2.5. Comparisonofcomponents
pieces of evidence pointto an origin in frontof the continuum
andbehindtheoutflow.Inbothcases,thevelocityoffsetstrongly
A summaryof the overallclassification scheme for the various
suggeststhatthecomponentsareassociatedwith J-typeshocks
componentsisshowninFig.6.Figure7thenshowstherelation-
(e.g.Hollenbach1997).
shipbetweenFWHM andbothvelocityoffsetandtheintensity
As already noted by Kristensen et al. (2013) for
inthe2 −1 (988GHz)transitionforthevariouscomponents
NGC1333-IRAS3A and Ser-SMM3, a few sources show spot 02 11
shockcomponentsinabsorption.Thesecomponentsaretoooff- scaledtothetypicaldistanceforthesampleof200pc.Asmost
setand/orbroadtobeconsistentwithabsorptionduetotheenve- ofthe outflowsfromthese sourcesarelargerthanthe Herschel
beamalongthe outflowaxis (see Yıldiz etal. 2014),the inten-
lopeorforegroundclouds.Inaddition,theyarepresentinexcited
sityofthecavityshockcomponentwascorrectedusingalinear
transitions which makes a foreground origin highly unlikely.
Off-positioncontaminationcanalsobeexcludedduetotheoffset scaling,i.e.assumingthattheemissionfillsthebeaminonedi-
fromthesourcevelocityandthatthe1 −1 position-switched rectionandispoint-likeperpendicularto it(Iobs(d/200)1).The
10 01
spot shockand quiescentenvelopecomponentsare assumed to
observationssharereferencepositionswithothersourceswhich
bepoint-like(I (d/200)2).
do not show these components. The depth of these absorption obs
features are consistent with absorption against the continuum Thoughthereareafewexceptions,thedifferentcomponents
only,suggestingthattheyoriginatebetweentheobserverandthe generallylieindistinctregionsoftheFWHMvs.offsetparame-
continuumsource,butnotbetweentheobserverandtheoutflow terspace,supportingtheideathattheyareformedunderdifferent
emission. conditions.In particular,the cavity shock and spotshock com-
We do not separate the “bullet” and less offset spot shocks ponentsarerelativelywellseparated.TheregionsofFWHMvs.
intoseparatecategoriesbecausetheinclinationoftheshockrel- offsetcoveredbythedifferentcomponentsfortheClass0andI
ativetothelineofsightplaysaroleinhowoffsetacomponent sourcesarealsosimilar.
is. However, in the cases where the offset from the source ve- The spot shock componentwhich lies in the middle of the
locityissmall,thespotshockcomponentsarealwaysnarrower cluster cavity shocks in the Class 0 FWHM vs. offset plot is
thanthecavityshock. thebroaderofthetwospotshockstowardsNGC1333-IRAS4A,
Thesuggestedphysicallocationofthespotshocks,whether markedwithablackarrowinFig.7.Thisislikelyrelatedtobow-
in the jet or at the base of the outflow, is indicated in shockswhichliewithintheHIFIbeamforthelower-frequency
Fig. 5. Bullets are observed to be small (few arcseconds) and transitions(seeAppendixB.3formoredetails).
A21,page9of49
A&A572,A21(2014)
102 resultsandanalysis,includingcomparisonwithotherresultsin
)
−1 Envelope theliterature,willbepresentedinSect.5.
s Cavityshock
m
Spotshock
(k 101 4.1.Integratedintensityratios
|
ource Aoffithrestwstaetper-inemsittutdinyginggasthieneyxocuitnagtiopnroatonsdtaprhsyissictoalucnodnedristtiaonnds
vs 100
theopacityoftheobservedtransitions,forwhichtherearefour
−
regimes. Below a certain N , a given transition will be opti-
vpeak 10−1 Class0 ClassI callythinwhileathighcolumH2nOdensityitwillbeopticallythick.
Bothofthesecasescanbeeitherinlocalthermodynamicalequi-
|
) 102 tlirbanrisuimtion(L,ToEr)s,ubw-htheenrmnHa2llyisexabciotveeditfhencri(cid:8)ticanl de.nAssitywafoterrthhaast
−1s largeEinsteinAcoefficientsandhighcriHt2icaldecnristities,thereisa
m
k 101 significantpartofrealisticparameterspacethatisopticallythick
K butsub-thermallyexcited.Inthisregime,thelinesaresaidtobe
](1 effectivelythinbecausethechanceofcollisionalde-excitationis
11 100 low,sophotonseffectivelyscatterwithintheregionandwillall
− eventuallyescapetheτ = 1surface.Assuch,the intensitystill
2
20 scalesasN ×n asintheopticallythinsub-criticalcase(see
I[ 10−1 Class0 ClassI e.g.LinkeeHt2Oal.19H727)eventhoughτ>1.
100 101 100 101 As discussed in Sect. 3.1, for those few sources and tran-
FWHM(kms−1) FWHM(kms−1) sitions where we can obtain H2O/H128O ratios, these suggest
that those components detected in H18O are optically thick in
Fig.7. Scatter plots of FWHMvs. offset of the peak from the source thosetransitions.However,thenumbe2roflines,componentsand
velocity(top)andintensityinthe202−111 linecorrectedtoacommon sources where this is the case is small. For sources or compo-
distance of 200pc (bottom) for the Gaussian components for Class 0 nentsforwhichH18Odataarenotavailableordetected,wecan
(left)andI(right)sources.Whenscalingtheintensities,alinearscaling alsousetheratioso2ftheintegratedintensityofthedifferentcom-
was used for the cavity shock components while a point-source scal-
ponentsin pairsofH16O lineswhichsharea commonlevel.In
ingwasused forthespot shock andenvelope components. Theblack 2
thelimitwherebothlinesareopticallythin,inLTEandhavethe
arrowindicatesthebroaderofthetwoshockspotstowardsNGC1333-
IRAS4AwhichisdiscussedfurtherinAppendixB.3. samebeamsize,followingGoldsmith&Langer(1999),theline
ratiobecomes:
I g A ν2
1 = u1 ul1 2 e(Eu2−Eu1)/kbTex, (1)
In general, the intensity of the components in the I g A ν2
2 u2 ul2 1
ClassIsourcesislowerthanfortheClass0s.Table4showsthe
numberofClass 0 andIsourcesinwhichthe cavityshockand where,foreachtransition,gu1isthestatisticalweightoftheup-
spotshockcomponentsaredetectedforeachtransition,aswell perlevel, Aul is theEinsteinA coefficientbetweenthe twolev-
asthemeanandstandarddeviationinthefractionalintensityin els, ν is the frequency, Eu is the upper level energy and Tex is
eachcomponentwithrespecttothetotalobservedintensity.For the excitation temperature.Alternatively,if both lines are opti-
thequiescentenvelopecomponentasthiscansometimesinclude callythick,inLTEandhavethesamebeamsizethelineratiois
bothabsorptionandemission,thiswascalculatedbysubtracting givenby:
theintensityoftheotherdetectedcomponentsfromthetotalob-
servedintensity,butmayincludeemissionandabsorptionwhich
canceleachotherout.Absorptionsinsomecomponentscanlead I1 = ν1(ehν2/kbTex −1)· (2)
toothercomponentshavinglargerintensitiesthanthetotal. I2 ν2(ehν1/kbTex −1)
While there is significant overlap in the intensity of com- Ifonelineisopticallythickbuttheotherisopticallythin,and/or
ponents in the lower panels of Fig. 7, the results in Table 4
ifthetransitionsaresub-thermallyexcited,thenthelineratiocan
showthatforagivensource,thecavityshockdominatesallthe
takearangeofvaluesdependingontheexcitationconditionsof
lines observed with HIFI, consisting of between 70 and 100%
thegas.
oftheintegratedemission.Thespotshockscontribute∼20%for
Figure8showssuchacomparison,coveringthemiddleand
Class0sourcesandareonaveragenegligibleforClassIsources.
upper excitation range probed by the water transitions acces-
The detection fraction of spot shocks is also much lower for
sible to HIFI. The intensity ratios for all components detected
Class I sources.Thequiescentenvelopedoesnothavea strong
in both lines are consistent with or close to the limit where all
contribution in the excited lines for Class 0 sources, though it
linesareopticallythick.Theopticallythinlimitshavebeencal-
can reduce the integrated intensity in the ground-state lines by
culatedforeachratioassumingexcitationtemperaturesof100,
upto20%dependingonthebalanceofemissionandabsorption.
300 and 500 K, to show that the temperature variation of this
It plays a more significant role in Class I sources, contributing
limitdoesnotimpactthe resultof thissimple analysis. For the
upto30%ofthetotalintensity. 3 −3 /3 −2 ratio, the lines come from the same upperen-
12 03 12 21
ergylevel,sotheopticallythinLTEratioisnotsensitivetotem-
perature.Whatismore,a searchofa wideparameterspaceus-
4. Analysis radex
ing the non-LTE molecular line radiative transfer code
In thissection we presentanalysis buildingon the results from (vanderTaketal.2007,discussedinmoredetailinSect.4.4)no
theprevioussection.Discussionofthewiderimplicationofthe non-LTE optically thin solutions where the 3 −3 /3 −2 is
12 03 12 21
A21,page10of49