Imaging galactic diffuse clouds: CO emission, reddening and turbulent flow in the gas around Zeta Oph

0.6 MB·English

by  H. S. Liszt

Astronomy&Astrophysicsmanuscriptno.10905-xar (cid:13)c ESO2009
January26,2009
Imaging galactic diffuse clouds:
CO emission, reddening and turbulent flow in the gas around ζ Oph
H.S.Liszt1,J.Pety2,3,andK.Tachihara4
1 NationalRadioAstronomyObservatory,520EdgemontRoad,Charlottesville,VA,USA22903-2475
2 InstitutdeRadioastronomieMillime´trique,300RuedelaPiscine,F-38406SaintMartind’He`res,France
3 Obs.deParis,61av.del’Observatoire,75014,Paris,France
4 NationalAstronomicalObservatoryofJapan,2-21-1,Osawa,Mitaka,Tokyo181-8588,Japan
receivedJanuary26,2009
9 ABSTRACT
0
0 Context.Mostdiffusecloudsareonlyknownaskinematicfeaturesinabsorptionspectra,butthosewithappreciableH contentmay
2
2 bevisibleintheemissionofsuchsmallmoleculesasCH,OH,andCO.
Aims.Weinterpretingreaterdetailtheextensiveobservationsof12COemissionfromdiffusegasseenaroundthearchetypicallineof
n
sighttoζOph.
a
Methods.The12COemissionisimagedinpositionandposition-velocityspace,analyzedstatistically,andthencomparedwithmaps
J
8 oftotalreddeningE∞B−VandwithmodelsoftheC+-COtransitioninH2-bearingdiffuseclouds.
Results. Aroundζ Oph,12COemissionappearsintwodistinctintervalsofreddeningcenterednearE∞ ≈0.4and0.65mag,of
B−V
which<∼0.2magisbackgroundmaterial.Withineitherinterval,theintegrated12COintensityvariesupto6-12Kkms−1,comparedto
]
A 1.5Kkms−1towardζOph.Nearly80%oftheindividualprofileshavevelocitydispersionsσv<0.6kms−1,whicharesubsonicatthe
kinetictemperaturederivedfromH towardζOph,55K.Partlyasaresult,12COemissionexposestheinternal,turbulent,supersonic
G 2
(1-3kms−1)gasflowswithespecialclarityinthecoresofstronglines.Theflowsaremanifestedasresolvedvelocitygradientsin
. narrow,subsonically-broadenedlinecores.
h
p Conclusions.ThescatterbetweenN(CO)andEB−Vinglobal,COabsorptionlinesurveystowardbrightstarsispresentinthegasseen
aroundζOph,reflectingtheextremesensitivityofN(12CO)toambientconditions.Thetwo-componentnatureoftheopticalabsorption
-
o towardζ Ophiscoincidentalandthestarisoccultedbyasinglebodyofgaswithacomplexinternalstructure,notbytwodistinct
r clouds.Theverybright12COlinesindiffusegasariseatN(H2)≈ 1021 cm−2 inregionsofmodestdensityn(H)≈ 200−500cm−3
t andsomewhatmorecompleteC+-COconversion.Giventhevarietyofstructureintheforegroundgas,itisapparentthatonlylarge
s
a surveysofabsorptionsightlinescanhopetocapturetheintrinsicbehaviorofdiffusegas.
[
Keywords. interstellarmedium–molecules

1
v
1
1. Introduction L204seenseveraldegreestothegalacticSouth.L204isclearly
1
outlined against the Hα emission from the ionized gas in the
1 Thelineofsighttothenearby(140-160pc)runawayO9.5Vstar star’sHIIregion(Gaustadetal.,2001).Inthiswork,the12CO
1
HD149757,ζ Oph,hasservedasthearchetypefordetailedob-
datacube from Tachihara et al. (2000) is employed to study the
1. servationalstudiesoftheinternalcompositionofdiffuse(AV<∼1 diffusegasatA ≈1magseennearerthestar.Wescrutinizethe
0 mag) clouds (Herbig, 1968; Morton, 1975), for optical/uv de- V
entireCOimageoftheabsorption-linehostwhoseoverallprop-
9 tectionofnewmoleculesindiffusegas(Maieretal.,2001)and
ertieshavesooftenbeeninferredfromonemicroscopicabsorp-
0 fortheoreticalmodelsofmoleculargasindiffuseclouds(Black
tionsightlinetowardthestar,andweinquiretowhatextentthat
:
v & Dalgarno, 1977; Van Dishoeck & Black, 1986, 1988; Kopp lineofsightfaithfullyrepresentsthehostgas.Moreover,large-
i etal.,1996).TheH2-bearingportionsofthegasoccultingζOph scalemapsofreddeningandextinctionhavebecomeavailableat
X
aredenseenoughtohostappreciablecolumndensitiesofcarbon
comparableresolution(thoughonlyalongtheentirelineofsight,
r monoxide,N(12CO)≈2.4×1015 cm−2(Morton,1975;Wannier
seeSchlegeletal.(1998)andDobashietal.(2005)),andweem-
a
et al., 1982; Lambert et al., 1994; Sonnentrucker et al., 2007), ploythesetocontrolagainstpossibleconfusionbetweendiffuse
and these are readily detectable in mm-wave emission toward
and darker sightlines, a source of concern given the strong CO
thestar(Knapp&Jura,1976;Liszt,1979;Langeretal.,1987).
lineswesee.
They were very partially mapped in CO emission (Kopp et al.,
The plan of this work is as follows. Section 2 summarizes
1996;Liszt,1997),aswellasCHandOH(Crutcher,1979;Liszt,
whatisknownobservationallyofthelineofsighttowardthestar
1997). and describes the pre-existing H I, CO, A∞ and E∞ datasets
COJ=1-0emissionaroundζOphwasimagedinmuchmore whicharediscussedhere.Section3discussVestheapBp−eVaranceof
complete fashion by Tachihara et al. (2000), who focused their theskyaroundζ Ophintermsofthestatisticsof12COemission
discussion on the properties of the nearby dark cloud complex andreddening.Section4discussesCOprofilesandlinewidthsin
termsoftheturbulentflowswhichareprominentintheemission
Sendoffprintrequeststo:H.S.Liszt profiles. Section 5 discusses physical conditions in the CO and
Correspondenceto:[email protected] H -bearing host gas, especially the regions of extremely bright
2
2 Liszt,PetyandTachihara:CO,reddeningandturbulencearoundζOph
Fig.1. Limiting reddening (Schlegel et al., 1998) and 12CO J=1-0 emission (Tachihara et al., 2000) in the vicinity of ζ Oph. The
12COemissioncontoursareshownatlevels1,2,4,6,8,10,12,16,20,24Kkms−1andhavebeencalculatedoverdifferentvelocity
rangesintheupperrightandlowerleftportionsofthemap,asindicatedatthemapcorners.ThedarkcloudcomplextotheSouth
ofζ OphisknownasL204andthegasattheupperright,aroundζ Oph,asL121.
(11-12K)COemission.Sect.6isasummaryandSect.7(avail- 2. Observations
able online) discusses the relationship between the extinction
and reddening measurements over the ζ Oph field and presents 2.1. Carbonmonoxide
someadditionalviewsofthe12COobservations.
ThedatacubeofTachiharaetal.(2000)comprisesnearly11,000
spectrafromtheNANTENtelescopewithabeamwidthHPBW
=2.7(cid:48) ona4(cid:48) gridingalacticcoordinates.Thespectrahave0.1
kms−1 resolution and the single-channel rms at this resolution,
0.5 K, is relatively high compared to that in the small numbers
Liszt,PetyandTachihara:CO,reddeningandturbulencearoundζOph 3
islikelyhostedinatomicgasoverlongpaths,asopposedtothe
more-localizeddiffuseanddarkcloudsofinteresthere.
A comparison between the limiting reddening of Schlegel
et al. (1998) and maps of the limiting extinction A∞ computed
V
fromstarcountsbyDobashietal.(2005)canbefoundinSect.
7(onlineonly).
2.4. Somegeneralconditionsinthegasalongthelineof
sighttoζ Oph
Infrontofthestar,E =0.32mag,N(HI)=5.2×1020cm−2,
B−V
N(H )=4.5×1020cm−2(Savageetal.,1977),N(12CO)≈2.4×
2
1015 cm−2 (Morton, 1975; Wannier et al., 1982; Sonnentrucker
etal.,2007),N(C0)=N(12CO)(Morton,1975)andN(C+)>∼3×
1017 cm−2 (Cardellietal.,1993).Themeankinetictemperature
ofthemoleculargasinferredfromH absorptionis54K(Savage
2
etal.,1977).
Toward the star the limiting extinction from the work of
Schlegeletal.(1998)isE∞ =0.55mag,sothatthebackground
B−V
reddeningisapproximately0.23mag.
The distance to the occulting material is generally taken to
be very close to that of the star, just outside the nearer edge of
Fig.2. Integrated H I line brightness from the LDSS survey the star’s H II region (Wood et al., 2005). At a distance of 140
of Hartmann & Burton (1997) at 35(cid:48) resolution, converted to pc,1(cid:48)correspondsto0.041pcand1oto2.44pc.
equivalent reddening E = N(H I)/5.8 × 1021 cm−2, N(H I)
(cid:82) B−V
=1.823×1018cm−2 T (HIdv)/(Kkms−1).
B
3. Theskyaroundζ Ophviewedinreddeningand
COemission
ofdemonstrationspectratypicallyshowninearlierwork(Liszt, Figure 1 is a composite image of the limiting reddening E∞
1997).StatisicallysignificantdetectionsoftheCOrequireW B−V
CO andintegrated12COJ=1-0emission,anupdatedversionofFig.
>∼1kms−1,whereWCOistheintegratedintensity 1 in Tachihara et al. (2000). The gray-scale underlay is the
Toward the star, we show the profile of Liszt (1997) from
limiting reddening (Schlegel et al., 1998) normalized to white
the then-NRAO Kitt Peak 12m telescope at 1(cid:48) (HPBW) spatial at the minimum value seen over the region, E∞ = 0.23 mag.
resolutionand0.12kms−1 spectralresolution(seeSect.3).To B−V
Superposedonthereddeningmap,theintegrated12COintensity
ensurethatthisprofileiscompatiblewiththosefromNANTEN
W hasbeencalculatedseparatelyoverthediffuse/translucent
werecentlyusedtheARO12mKittPeaktelescopetore-observe CO
northwest region, referred to as L121, and the translucent/dark
several positions having comparatively strong emission in the
southeast region, L204. For the diffuse gas of L121 at upper
NANTENdatacube.The12mspectraagreewiththeNANTEN
right,theredandbluecontourscorrespondtothevelocityranges
datatobetterthan5%,aremarkablecoincidenceconsideringthe
differenceinhardwareandspatialresolution. aboveandbelowv=-0.25kms−1asindicatedintheupperright
corner: this division corresponds to the natural separation be-
tweenthetwocomponentsofCO,CHandOHemissionfound
2.2. HI towardandaroundζ Oph,asshownintheprofilesandposition-
velocitymapsofLiszt(1997).ForthedarkergasofL204tothe
To help distinguish between foreground and background mate- southeastW wascalculatedoverthreeintervalsbutnearlyall
CO
rial, or atomic and molecular gases, we employed the H I pro- of the emission from L204 occurs in the velocity interval 2-6
files from the Leiden-Dwingeloo all-sky H I survey (Hartmann kms−1representedbythewhitecontours(seeFig.5).Thenoise
&Burton,1997).Thesedatahave35(cid:48) resolutionona0.5o grid level is somewhat larger for these contours, because W was
CO
ingalacticcoordinates. calculatedoverasomewhatbroaderinterval.
ThestrongCOemissionassociatedwithL204oftenfollows
theridgelinesoftheextinctionwithsomethingofasetback(red-
2.3. Reddeningandextinction
dening without apparent CO) in the direction of the star. This
The reddening toward ζ Oph is known to be E = 0.32 mag is consistent with the edge-on geometry for L204 described by
B−V
(Morton,1975),butmapsoftheforegroundextinctionoveronly Tachihara et al. (2000). Nearly all of the CO emission in the
thefirst140pcareunavailable.Instead,weemploythelimiting L204 dark cloud complex is found in the interval v = 2-6
lsr
reddening(fromheretoinfinity)fromtheworkofSchlegeletal. kms−1 and so does not overlap that of the more diffuse gas to
(1998),denotedbyE∞B−V,withaspatialresolutionof6.1(cid:48),pub- the north, seen at vlsr <∼ 2 kms−1. The extent to which the two
lished on a 2.5(cid:48) grid; the stated global rms error of this dataset regionsareatrestinthedirectionsjoiningthem,andmightpar-
is 16% (a percentage at each pixel). Their values are based on takeoftheopticalpumpingexcitationmechanismfordiffusegas
a determination of the dust column density estimated from the describedbyWannieretal.(1997)isunknown.Tachiharaetal.
IRAS 100 micron flux adopting the temperature variation de- (2000)assumedthatthetwocloudswereco-movingattheedge
rivedfromCOBE/DIRBE240microndata.Theminimumlim- oftheHIIregionaroundζOphinordertodiscusstheenergetics
itingreddeningintheregion,approximately0.23mag(Fig.1), ofthegas.
4 Liszt,PetyandTachihara:CO,reddeningandturbulencearoundζOph
which include all of the unrelated background and foreground
gas.Immediatelyaroundζ OphE∞ =0.55mag,comparedto
B−V
E = 0.32 mag in front of the star, so some 0.55-0.32 mag
B−V
= 0.23 mag of reddening occurs behind the star. This also cor-
responds well to the absolute minimum in Fig. 1, E∞ = 0.23
B−V
magfoundsome1.5oNorthofthestar.
To estimate the amount of extraneous foreground material,
wenotethatthisisexpectedtobeinatomicformand,inabsorp-
tionagainstthestar,N(HI)= 5.25×1020 (Savageetal.,1977)
correspondingtoEB−V<∼0.1mag.ThisleavesEB−V>∼0.22mag
of foreground material associated with the pure H component
2
and gives the general idea that perhaps as much as 0.32 mag
should be subtracted from the map of E∞ to infer the local
B−V
reddeningintrinsictotheH -bearinggas.However,thisisprob-
2
ablyanoverestimateoftherequiredcorrectionbecausesomeof
the foreground atomic gas is in the vicinity of L121 and L204,
providingshielding.
Figure 2 is a map of the integrated H I brightness,
scaled to the mean relationship between hydrogen column
density and extinction, that is, the map shows 1.823 ×
(cid:82)
1018 cm−2 T (HI)dv/5.8×1021 cm−2mag−1 where the veloc-
B
ityintegralisinunitsofKkms−1.Severalofthelocalreddening
peakswhichlackCOemissioninFig.1evenatE∞ =0.65mag
B−V
arepresentaspeaksinFig.2,forinstanceat(l,b)=(7.5o,21.5o)
Fig.3.IntegratedintensityoftheCOlineW vs.limitingred-
deningE∞ .Top:forthediffuseregionL12C1Oatupperrightin and(4o,20o).Someofthisgasmustbeindigenoustotheregion
B−V of interest, but the extinction associated with the H I peaks is
Fig. 1; bottom: for the darker L204 gas at lower left in Fig. 1.
only ≈ 0.06 mag, judging from the peak levels in Fig. 2 (0.28
Ineachcasethedataareshowntwice,atrightasindividualdata
mag)comparedtothenearbybackgroundlevel(0.22mag).
points, at left binned and contoured at logarithmic (factor two)
TheminimainFig.1andFig.2andthedifferencebetween
intervals: the contour representation of the L121 data is super-
the foreground and limiting reddening toward the star consis-
posed at bottom. Also, in each case the data are divided into
tentlyimplyabackgroundreddeningcontributionof≈ 0.2mag
regions of lower and higher extinction with the separation oc-
curingatthemeanE∞ overdatapointswhichlackstatistically overtheregionofinterest.
B−V
significantdetectionsofCOemission(0.46magand0.67magat
topandbottom,respectively).Thelocusofthesightlinetoward
3.2. QuantitativerelationshipbetweenCOemissionand
ζ OphintheB-regionoftheupperpanelisindicatedbya(red)
reddening
cross.Mean12COemissionprofilesareshowninFig.5.
To quantify the relationship between reddening and CO emis-
sion,andtocompareandcontrastthediffuseanddarksightlines
TotheNorthofL204,oneithersideofthestaristheζ Oph we divided the extent of Fig. 1 along a Northeast-Southwest
diffusecloud,describedastheL121complexbyTachiharaetal. diagonal in the trough of reddening between L121 and L204,
(2000).Mostoftheemissionfromthatgasoccursat-2.4kms−1 alongalinerunningfrom(l,b)=(3o,19.333o)to(10o,25o).Fig.
<vlsr<2.2kms−1.Thered-shiftedkinematiccomponentat-0.2 3showstheintegratedCOintensitiesWCO andlimitingredden-
kms−1 <v <2.2kms−1whichisseeninCOemissionandin ing E∞ for all points in both regions: the profile integral was
lsr B−V
many species in absorption toward the star, is very much con- takenovertherange-2.4kms−1 <v <2.2kms−1 forthedif-
lsr
finedtothenorthernandnorthwesternedgesofthebroaderdis- fuse L121 gas shown in the top panel, and 2 kms−1< v < 8
lsr
tribution of blue-shifted gas. Its separate identity as a second kms−1 for the darker L204 region in the lower portion of the
cloud is somewhat marginal in the emission maps and hardly map of Fig. 1 and the lower panel in Fig. 3. The rms noise in
supported by a more detailed examination of the turbulent gas integratedintensityis0.5-0.6Kkms−1.Profileintegralsabove1
kinematics(alsoseeFig.7). Kkms−1generallyrepresentrealdetections.
The diffuse gas in the L121 complex around ζ Oph is sep- OncewerealizedthattheCOemissioninthediffuseregion
arated from the L204 dark cloud by an extended trough in wasbimodal,asillustratedinthetoppanelsofFig.3,wefurther
the reddening whose overall mean value is E∞ ≈ 0.43 mag. sub-dividedthediffusegasintoAandBportionscorresponding
B−V
Immediately below the star is a pronounced minimum whose to the two branches of the emission distribution. The A and B
meanisE∞ ≈0.34magandwhoseabsoluteminimumisE∞ portionswereseparatedatE∞ =0.455mag,whichisthemean
=0.29maBg−.VTheoverallimpressionisofacylindricalshellBg−eV- E∞ for those sightlines alBo−nVg which W < 1 K kms−1and
ometry and perhaps a separate, more circular shell of radius ≈ wBh−icVh therefore lack statistically significaCnOt detections of CO
1o around the star. However, a yet-larger map of the extinction emission.TheA-branchpixelshavestrongCOemissionatE∞
showsthatL204ispartofamuchlargerridge. substantially below the mean of those sightlines lacking BC−OV
emission at all. For the diffuse gas in L121 there is actually
a substantial spatial segregation of the A and B portion pixels,
3.1. Separatingforegroundandbackgroundreddening
with unweighted mean < (l,b) > = (6.0o±0.7o,23.8o±0.7o) and
Asnotedabove,wewishtounderstandtherelationshipsbetween <(l,b)>=(4.0o±0.9o,22.3o±1.0o)fortheAandBportions,re-
W and reddening, but we only have maps of the reddening spectively. These centroids are on opposite sides of ζ Oph and
CO
Liszt,PetyandTachihara:CO,reddeningandturbulencearoundζOph 5
3.3. Statisticsofthelinebrightness
Figure4presentsthedistributionofintegratedandpeakbright-
ness. The differences between the A (lower E∞ ) and B por-
B−V
tions of the diffuse gas are somewhat more pronounced in the
distributionofthelineprofileintegralatleftandsomewhatless
so in the peak, so that the line widths differ more than the line
heights.Althoughthemedianbrightnessisatmostonly3Kor3
Kkms−1inthediffusegasandthetailofthedistributionseems
very poorly populated above, say, 8 K peak brightness: paths
which traverse L121 even in the short dimension have a very
substantialchanceofcontainingatleastonesuchbrightline,as
discussedinSect.5(seeFig.7).
It is not the intention here to discuss the dark gas, but it
shouldbenotedthatthedifferenceinmeanbrightnessbetween
thediffuseanddarkregionsaremodestandcorrespondapprox-
imately to the differences in E∞ , thereby preserving the pos-
B−V
sibilityofacommonCO-H conversionfactor;thesameisalso
2
true for the A and B portions of L121, see Fig. 5. This occurs
despitethefactthatmostofthefreegas-phasecarbonisinC+in
L121 (99% toward ζ Oph) and in CO in the dark gas of L204
(whereN(CO)≈3×1017cm−2)implyingadifferenceinCOcol-
umndensityandW /N(CO)byafactoroforder50.InSection
CO
5 we discuss the very different proportionality W ∝ N(CO)
CO
Fig.4.Cumulativeprobabilitydistributionsoftheintegratedin- whichisobservedwithinthediffuseregimealone(Liszt,2007a).
tensityW (left)andpeaklinetemperature(right)overthedif-
CO
fuse,northeasternL121(attop)anddarksoutheastL204sight-
3.4. AnincidentalboundonN(H )overthediffusegas
linesshowninFig.1.OnlythosesightlinesatwhichW ≥1K 2
CO
kms−1havebeenconsidered. Becausethelineofsighttoζ Ophoccursatsuchahighvalueof
E∞ relativetotherestofL121weinferthatN(H )isneververy
B−V 2
muchlargerinL121thantowardthestar.Forinstance,ifwetake
E∞ = 0.65 mag characteristic of the B-portion and subtract a
separated by more than 1σ+1σ in each coordinate. The red- baBc−kVgroundcontribution0.23magequaltothattowardthestar,
shifted gas appears mostly in the higher-extinction B-portion theremaininggascolumnwithE∞ =0.65-0.23=0.42mag
B−V
while the blue-shifted gas appears more nearly in both the A corresponds to N(H) ≈ 2.4 × 1021 H-nuclei cm−2 and N(H )
2
andB-portions. < 1.2×1021 cm−2.Conversely,becausethelineofsighttoward
In either emission branch, WCO varies widely over a rela- the star has such a high value E∞ = 0.55 mag relative to the
tivelynarrowrangeinE∞ ,mimicingtheextremesensitivities rest of the region, emission fromB−Vthe A-branch at E∞ = 0.4
B−V B−V
ofN(CO)whichareseeninglobalabsorptionlinesurveysofsin- mag,probablyarisesinregionswhosereddeningandN(H )are
2
glesightlinesoverwidely-separatedregions(seeLiszt(2007a)). actuallybelowthoseseentowardthestar.
Thewidthsofthetwobranchesarecomparabletothestatedrms
noiseoftheE∞ data(16%)andtheyareseparatedby0.2mag,
B−V
which is of the same order as the background contribution: it 4. Lineprofiles,linewidthsandturbulentflowsin
ispossiblethattheywouldmorenearlycoincideinEB−V ifthe diffusegas
background contribution were strongly variable. However, the
B-branchathigherE∞ ismoreheavilypopulatedatW >5 4.1. COlineprofilesandprofilewidths
B−V CO
Kkms−1,suggestingthatitactuallyissomewhatmorestrongly For a pure-H gas, the Doppler temperature equivalent to a
2
shielded,fosteringahigherCOabundanceandbrightness:inthis given linewidth is T = 43.6 K FWHM2 or T = 242 K
dopp dopp
regime,WCO∝N(CO)(ibidandseeSect.6)andtheconversion σ 2 where σ is the velocity disperson in kms−1. Typical ki-
offreecarbonfromC+toCOoccursoveraverynarrowinterval nevtic temperavtures in CO-bearing diffuse gas are above 30 K
inE and/orN(H ).
B−V 2 (Sonnentrucker et al., 2007; Burgh et al., 2007; Liszt, 2007a)
The distribution of brightness in the darker L204 region is andthemeankinetictemperatureseeninsurveysofH absorp-
2
not similarly bimodal at all E∞ and there is less rationale for tionis70-80K(Rachfordetal.,2002;Savageetal.,1977).CO
B−V
asimpledivisionintosub-portions.Howeverwhenthisisdone, profiles with FWHM <∼ 1 kms−1 are subsonic at diffuse cloud
atE∞ =0.67magcorrespondingagaintothemeanoverpixels temperatures and the purely-thermal velocity dispersions of
B−V
lackingstatisticallysignificantCOemission,theresultantlower- COmolecules,0.1kms−1atT =50K,donotcontributeim-
K
E∞B−V portioncorresponds(inE∞B−V andWCO)totheentiretyof portantlytotheobservedlinewidths.
thediffuseregion.Consequentlythelower-E∞ partofL204is Unweighted mean profiles formed over the sub-portions of
B−V
labelled L204 AB and the other L204 C. The behavior of WCO thediffuseL121anddarkL204regionsareshowninFig.5.The
with E∞ in the dark gas is complex but clearly bimodal for meanprofileshavelinewidthswhicharesupersonic,FWHMof
B−V
WCO >∼12Kkms−1;thestrongestemissionisbynomeanslim- typically 2-3 kms−1 but the individual sightlines in L121 typi-
ited to the darkest regions. Statistics of the brightness distribu- callyhavesubsoniclinewidths;theFWHMofthetwokinematic
tion over L121 and L204 are shown in Fig. 4 and discussed in componentsseentowardthestarat1(cid:48) resolution(showninFig.
thefollowingsections. 5asthedarkdashedline)are0.6kms−1and1.1kms−1,equiva-
6 Liszt,PetyandTachihara:CO,reddeningandturbulencearoundζOph
Fig.5. Mean spectra over the diffuse and dark sub-portions de- Fig.6. Statistics of line profile velocity dispersions in the dif-
finedinFig.3:theprofilesinthediffusegas(L121)arethoseat fuse gas. Large outer panel: cumulative probability distribution
lower velocity. The integrated brightnesses are W = 2.9 and histograms of the measured velocity dispersion σ shown sep-
CO v
4.0Kkms−1 forL121and2.3and7.4Kkms−1 forL204.Also arately for the A and B portions of the diffuse gas (see Sect. 4
shown(dashedblackline)isaspectrumat1(cid:48) resolutiontoward of the text). The one-dimensional thermal velocity dispersions
ζ Ophfromthe12mtelescope. correspondingtokinetictemperaturesof20,40and80Kinagas
ofpureH areshownatupperleft.Insetatlowerright:variation
2
ofσ withW .
v CO
lenttoDopplertemperaturesofonly16Kand53KinapureH
2
gas,comparedtoameasuredtemperatureinH of54Kasnoted
2
inSect.2.4. gradientsinthehostmedium,i.e.thecharacteroftheturbulence.
ThedistributionofvelocitydispersionsfoundovertheL121 Althoughveryelaborateanalysesoflineprofilecentroids(Pety
region is shown in Fig. 6. To produce this figure we used the & Falgarone, 2003) and fluctuations in the wings of optically
followingwindowingtechniquetomeasurethewidthsofspec- thickprofiles(Falgarone&Phillips,1996)havebeenusedtoin-
tra which might contain more than one kinematic component: ferthepropertiesofturbulenceindenserneutralmedia,turbulent
select the overall velocity interval of the diffuse gas; find the flowsinthenearbydiffusegasofL121areimmediatelyvisible
peakchannel;selectthosecontiguouschannelsaroundthispeak inthelinecoresand(usually)spatiallyresolvedintotheshiftsof
with temperatures above a noise threshold of 0.25 K; calculate individual,subsonicprofiles.
thebrightness-weightedvelocitydispersionoverthosechannels; ThisisillustratedforζOphinFig.7,whereweshowacom-
maskoffthatportionoftheprofile;repeatuntilnochannelabove prehensiveseriesoflongitude-velocitydiagramsspacedatone-
1 K remains unmasked. The dispersions measured en masse in pixel(4(cid:48))intervalsingalacticlatitude,aslabelledintheindivid-
thiswayaresubjecttooverestimatingthewidthincasesofun- ualpanels.Thepeaklinebrightnesses(K)seenintheindividual
recognizdblending,buttheyagreetowithin10%foraseriesof panelsofFig.7arelabelledonthebarsshowingthecolorscal-
test profiles which were fit with gaussian components (for in- ingineachpanel.Althoughlineswithpeakbrightnessesabove
stance,thoseshowninFig.8). 8KseemrelativelyrareinFig.4,theyarecommonenoughthat
The general properties of the gas in L121 have often been peak brightnesses 7.5 K and higher appear in 40%, 18 of the
inferredfromprofilesseenalongthesingle,microscopicabsorp- 44individualpanelsofFig.7.AsshowninFig.3,themostin-
tion line of sight toward the star and a sensitive CO profile to- tenselinesinthediffusegasarebynomeanslimitedtothemore
ward ζ Oph at 1(cid:48) spatial resolution is also included in Fig. 5. heavily-reddenedsightlines.
How representative is it of the larger-scale distributions of the Figure7showsthatadescriptionofthegasintermsoftwo
host gas? The stronger CO component at vlsr = -0.7 kms−1 to- identifiableforegroundcloudsattheabsorptionlineorCOemis-
ward ζ Oph is one of the narrowest lines known in a diffuse sionlinevelocitiestowardζ Ophisnotappropriate.Thelineof
cloud, with FWHM = 0.60 kms−1. Broad consideration of this sight to the star could just as well have occured behind any of
questionisgiveninthediscussionofFig.7whichdisplaysmost theprofilesexhibitedinFig.7,leadingtoawidevarietyofpos-
oftheprofilesseenintheL121region,immediatelyfollowing. sible interpretations of the intervening medium. Discussing the
mapinFig.1,wesawthatthered-shiftedcomponentgenerally
appears as something of a fringe at the northern edges of more
4.2. Velocitygradientsandturbulentflows
broadly-distributed, negative-velocity gas. Viewed in position-
The connections between the profiles seen at individual pixels velocityspaceinFig.7,thered-shiftedgasisoftenseenasapro-
andthemeanprofilesshowninFig.5aretheflowsandvelocity nouncedkinematicexcursionorwing,forinstance,atb=22.8o
Liszt,PetyandTachihara:CO,reddeningandturbulencearoundζOph 7
or 23.4o. Further to the North, at b > 23.2o, the kinematic pat- includinggrain-assistedneutralizationofatomicions(including
ternundulatesacrossthepositive-velocityportion.Atb=23.4o mostimportantlytheprotons,ibid),toallowequilibriumofthe
theresolvedvelocitygradientspanstheentirerangeofvelocities self-shieldingof H formedon grainsandthe formationof CO
2
inthediffusegas:viewedfromadifferentperspective,thissame bythermalelectron-recombinationofafixedquantityofHCO+
regionmighthavebeenseenasasingle,broadline. X(HCO+)=N(HCO+)/N(H )=2×10−9(theactualsecularevo-
2
Figure 8 is an expanded view of the panel in Fig. 7 at b= lutionistracedbyLiszt(2007b));and,tocalculatetherotational
22.6o, where we also show several included line profiles, their excitationofCOassumingmicroturbulentradiativetransferwith
gaussiandecompositionandtheresultantFWHM.Exceptatthe a linewidth determined by the local sound speed. Note that the
centerofthediagramwherethereisapartiallyspatially-resolved interstellar radiation field in the models has not been increased
velocity gradient, the FWHM are small enough to be subsonic abovethemeaninterstellarvaluetoaccountforthepresenceof
for an H gas at typical diffuse cloud temperatures above 30 thestarandthatalloftheaspectsshownhereforthemodelre-
2
K.Thustheobservationsshowindetailhowsupersonicprofiles sultsarepresentintheabsorptionlinedatasummarizedbyLiszt
mightarisefromthecoincidentalsuperpositionoraddition(for (2007a).
instance through beam-smearing) of velocity-shifted quiescent Model results are shown for just two fairly high densities
regions. n(H)=256cm−3and512cm−3.ToformthegraphsinFig.9,re-
AsinFig.8therearemanyhighly-localized,relativelybroad sultswerederivedbyintegratingalongsightlinesrangingacross
linesin4(cid:48)pixels,oftenjoinedtoabruptbutspatially-resolvedve- thefacesofmodelspheresfromcentertoedge.Thegraphssum-
locitygradientsandreversalsinnarrow-linedgas.Thissuggests marizeresultsgleanedfrommodelswhosecentralcolumnden-
that broader lines are themselves composed of unresolved ve- sitiesN(H)variedwidely,sothatthesamevalueofN(H )might
2
locity gradients and it seems possible that any profile in Fig. 7 occuratdifferentpositionsacrossthefacesofmodelswithdif-
withawidthsubstantiallyabovesonicisanunresolvedgradient. fering n(H) and N(H), and therefore have different N(CO) and
For instance, compare the velocity span at b=22.87o, which is W .ForN(H )=1×1021 cm−2 themodelshavetypicalsizes
CO 2
spatiallyresolved,withthatattheedgeoftheemissionregionat of1.2-2.5pcbuttheCOabundanceisconcentratedintosmaller
b=23.33o.Recallingtheextremelynarrowblue-shiftedCOcom- centralportionsofthehostbodiesowingtothechemistryofCO
ponenttowardζ OphandthelowimpliedDopplertemperature, and that of H , and the CO emission is more concentrated still
2
wewonderwhetherandatwhatscaleprofileshavinglinewidths owingtogeometryandradiativepumping.
thataresubsonicatthehighertemperaturesexpectedfordiffuse As noted above N(H ) = 0.5 − 1 × 1021 cm−2 in the CO-
2
gas,say60K,mightshowvelocityorspatialsubstructure.The emittingregionsaroundthestar.Asshowninpanelcatthelower
geometryoftheturbulentflowsproducingthepatternsinFig.7 left,thisispreciselytheregimewherecarbonisabouttorecom-
willbeconsideredinasubsequentpaper. binefullytoCOatthequoteddensities:theincreaseofN(CO)
with N(H ) is very rapid. Both N(CO) and W vary rapidly
2 CO
and have large scatter when plotted against N(H ). Substantial
5. Physicalconditionsinthediffusegas 2
CO column densities can accumulate in gas which is still rela-
There are some 1000 sightlines with WCO >∼ 1 K kms−1 in tively warm, 30-50 K, giving rise to 12 K lines as observed in
the diffuse, Northwest portion of the map in Fig. 1. At a dis- thebrightestprofiles.
tance of 140 pc, the H -mass associated with these sightlines, The most important consideration is as shown at the upper
2
parametrized by their average H2 column density <N(H2)>, is rightinpanelb,WCO ∝N(CO)forWCO <∼10Kkms−1,explic-
M=430 M <N(H )/1×1021 cm−2 >wheretypicalvaluesof itlyindependentofdensityandimplicitlyindependentofN(H )
0 2 2
N(H )alongtheL121sightlinesare0.5−1.2×1021 cm−2 (see and other cloud properties. This is a very general consequence
2
Sect.2.4and3.4).Thismaybecomparedwiththevalue520M of very sub-thermal excitation, as first shown by Goldreich &
0
given by Tachihara et al. (2000) based on a CO-H conversion Kwan(1974)anddoesnotrequirelowopticaldepth.Asshown
2
factorN(H )/W =1.56×1020cm−2/(Kkms−1). in Fig. 4, some 80-90% of the diffuse cloud spectra have W
2 CO CO
Nonetheless, deriving the physical parameters of host dif- < 5-6 K kms−1 and virtually all are below 10-12 K kms−1,
fuse gas from CO profiles is challenging. Strong fractionation just in the regime characteristic of sightlines studied in uv and
of the carbon isotopes (Liszt & Lucas, 1998; Sonnentrucker mm-wave CO absorption work generally, see Fig. 6 of Liszt
et al., 2007; Burgh et al., 2007; Liszt, 2007a) causes the (2007a).Therefore,the12CObrightnessmapinFig.1isamap
N(12CO)/N(13CO)tovaryintherange15-150,thusmakingit of N(12CO) in the diffuse gas and the same would be true for
impossibletoderivethelineexcitationtemperaturesandoptical N(13CO) and the brightness of the 13CO line. This is the one
depths,orthekinetictemperaturesandcolumndensities,under unambiguousresultofmappingCOemissioninanydiffusegas.
the usual assumption (valid in dark gas) that the relative abun- TheextremesensitivityofW toN(H )thereforearisesbe-
CO 2
dances of 12CO and 13CO only reflect the general interstellar causeW ∝N(CO),sothataplotofW vs.N(H )isequiv-
CO CO 2
carbonisotoperatio.However,itisstraightforwardtoshowthat alenttoplottingN(CO)againstN(H ).Thenetresultisthatal-
2
thegeneralpropertiesoftheCOobservations,evenincludingthe though commonly-used values of the CO-H conversion factor
2
ratherunexpectedlybrightlines,fiteasilyintotheframeworkof apply to some gas (as indicated in panel a of Fig. 9) the CO-
diffusegasattypicaltemperatures30-60Kandmodestdensity. H conversionfactorvarieswidelyindiffusegasandtheactual
2
The real underlying mysteries are the working of the poorly- N(CO)/N(H ) ratio is small but very uncertain. A map of CO
2
understoodpolyatomicchemistrywhichformstheCOandother emissionfromdiffusegasisanimageofthechemistry,notthe
speciesatsuchmodestdensities(Lisztetal.,2006),andtheori- massdistribution.
ginoftheturbulentflowswhichmaypowerthechemistry. Last,notethattheJ=1-0COlinebrightnessisinsensitiveto
Figure 9 reports some results from the models which were densityatfixedN(CO),indicatingthatothertracersarerequired
usedtointerpretCOabsorptionlinedatabyLiszt(2007a).The to measure the local density when mm-wave emission profiles
underlying physics are: to heat a uniform-density gas sphere are analyzed. The J=2-1/J=1-0 line brightness ratios at lower
by the photoelectric effect on small grains, as in the work of right in Fig. 9 are not very sensitve to density, which explains
Wolfire et al. (1995, 2003), to calculate the ionization balance why line brightness ratios 0.7 - 0.75 are indeed so commonly
8 Liszt,PetyandTachihara:CO,reddeningandturbulencearoundζOph
Fig.7.Longitude-velocitydiagramstraversingtheζ OphdiffusecloudatmanylatitudeswithintheNorthwestportionoftheregion
showninFig.1.Theverticalseparationbetweenpanelsis4(cid:48) (1pixel)andthegalacticlatitudewithineachpanelislabelled.The
colorscaleineachpanelareseparatelynormalizedwithinarange(Kelvins)shownontheindividualcolorbars.
observed in diffuse and translucent gas (Falgarone et al., 1998; 6. Summary
Pety et al., 2008). The J=3-2/J=1-0 brightness ratio is a better
Thelineofsighttothenearby(140-160pc)runaway09.5Vstar
indicator of density in the CO lines, but care must be taken to
ζ Ophhasformanyyearsbeenusedasanarchetypeforstudy-
matchthespatialresolutionofthetwolines.
ingthepropertiesofdiffusecloudsinopticalanduvabsorption.
Because the material has an appreciable molecular content, the
host diffuse clouds can actually be imaged on the sky in space
and radial velocity. Because the gas is well extended and com-
parativelyclose,itprovidesanunusualopportunityforstudyof
Liszt,PetyandTachihara:CO,reddeningandturbulencearoundζOph 9
Fig.9. Models of 12CO formation and excitation. Results are
shown for models having n(H) = n(H I) + 2n(H ) = 256 cm−3
2
and 512 cm−3. Upper left: Variation of integrated 12CO bright-
Fig.8.Thelongitude-velocitydiagramatb=22.6o fromFig.7, ness WCO with N(H2); the shaded line labelled “2” is a CO-
H conversion factor 2 × 1020 H cm−2 (K kms−1)−1. Upper
expanded and decomposed into its constituent profiles at three 2 2
right: variation of W with N(12CO). Bottom left: variation
positions (shown inset). Overlaid on each inset spectrum is a CO
of N(12CO) with N(H ). Lower right: comparison of 12CO line
one-componentgaussianfitwhoseFWHMisshownattheupper 2
brightnessesinthelowesttworotationaltransitions.
left.
diffuse gas and its interaction with its surroundings, including flows in this gas are in general directly visible as the spatially
thestar. andkinematicallyresolvedvelocitygradientsinsimple,narrow,
We began by comparing the 12CO J=1-0 emission line dat- brightlineswhosewidths(σ <0.6kms−1)aresubsonicatdif-
v
acubeofTachiharaetal.(2000)(HPBW2.7(cid:48)observationsona4(cid:48) fuse cloud temperatures T = 30 - 60 K (Fig 4 and 8). Other,
K
=0.1pcpixelgrid)withamapofthelimitingreddeningE∞ locally-broaderlineprofileswilllikelyberesolvedintosuchve-
B−V
from the work of Schlegel et al. (1998) having similar resolu- locityshiftsofnarrowlinecoreswithhigher(than4(cid:48))resolution
tion6.1(cid:48) (Fig.1).ThereddeningintheL121complexnearand although this remains to be tested. Conversely, it also remains
aroundζ OphrangesfromE∞ =0.23-0.75magandtheinte- tobeseenwhetherCOlineprofileswhicharesubsonicbutstill
B−V
grated CO emission up to W = 12K kms−1, with 12 K peak supra-thermalat2.7(cid:48)resolutionwillshowspatialorvelocitysub-
CO
temperatures, which are very bright lines indeed. Comparison structurewhenmappedathigherresolution.
ofreddeningofthestar(0.32mag)andthroughtheMilkyWay Webrieflydiscussedsomemodellingresultsoftheformation
(E∞ = 0.55 mag) in the same direction, and comparison of H and excitation of CO in diffuse media (Fig. 9). The brightness
B−V
I seen toward the star in uv absorption and around the star in of the strongest CO lines can be understood by noting that the
21cmemission(Fig.2),suggeststhat≈0.2magofE∞ should H columndensitiesinthegasaroundζ Opharenearthepoint
B−V 2
beascribedtounrelatedbackgroundmaterial. (N(H ) ≈ 1021 cm−2) where carbon fully recombines to CO at
2
CO emission from diffuse L121 gas seen around ζ Oph is evenmodestdensitiesn(H)=200−500cm−3,sothatsubstantial
bimodal in E∞ , clustering around E∞ = 0.4 mag and 0.65 columnsofCOmayformingaswhichisattypicaldiffusecloud
B−V B−V
mag and varying widely (1 K kms−1 < W < 6-12 K kms−1) temperatures (above 30 K). In turn, such densities will excite
CO
withE∞ inoneoftwonarrowranges,seeFig.3:thesamelarge CO to the required degree even though they are far too low to
B−V
scatterinCOcolumndensitywithreddeningandN(H)whichis thermalizethelowerrotationallevelpopulations.
seen globally in galactic absorption line surveys also occurs in In the range WCO <∼ 10 K kms−1 it is the case that WCO ∝
the single region studied here. The lower-reddening branch of N(CO), as a consequence of very sub-thermal excitation. The
the emission is spatially segregated to the galactic northeast of CO-H conversion factor therefore varies widely in diffuse gas
2
the star and has somewhat smaller mean integrated brightness (becauseN(CO)variesrapidlywithN(H )andwithgreatscat-
2
W (Fig. 4) and velocity dispersion σ (0.35 kms−1 vs. 0.42 ter), but it takes on values N(H )/W = 2×1020H cm−2/(K
CO v 2 CO 2
kms−1;Fig.6).However,peaktemperatures8-12Karepresent kms−1)inlimitedcircumstances(Fig.9panelaatupperleft).
inbothbranches. If W ∝ N(CO), a sky map of W like that in Fig. 1 is
CO CO
Thetwomoststrikingobservationalresultsofthisstudyare amapoftheinterstellardiffusecloudchemistry.Thisshouldbe
thestronglineswhichemanatefromthediffusegas,uptoW contrastedwiththemoreusualassumptionofaconstantCOcon-
CO
=12Kkms−1,andthevelocitystructurepresentinthestrongly version factor N(H )/W , in which case a CO map traces the
2 CO
emitting 12CO line cores there (Figs. 7 and 8). The turbulent contoursofthemass(i.e.thebulkmolecularmaterial)largelyin-
10 Liszt,PetyandTachihara:CO,reddeningandturbulencearoundζOph
dependentofphysicalconditions.Eventhedarkergasseeninthe —.1997,A&A,322,962
L204 complex South of ζ Oph is not immune to the influences —.2007a,A&A,476,291
ofchemistry,whichareclearlyvisibleinthedisplacementsbe- —.2007b,A&A,461,205
tweenregionsofhigherW andE∞ . Liszt,H.S.&Lucas,R.1998,A&A,339,561
CO B−V Maier,J.P.,Lakin,N.M.,Walker,G.A.H.,&Bohlender,D.A.2001,ApJ,553,
The wealth of structure seen in the foreground CO bright- 267
ness map has important implications for absorption line study Morton,D.C.1975,ApJ,197,85
ofdiffuseclouds.Sincetheviewinggeometryisanaccidentof Pety,J.&Falgarone,E.2003,A&A,412,417
Pety,J.,Lucas,R.,&Liszt,H.2008,A&A,inpress
the relative locations of the Sun and ζ Oph, our line of sight
Rachford,B.L.,Snow,T.P.,Tumlinson,J.,Shull,J.M.,Blair,W.P.,Ferlet,
to the star could equally well have intercepted any of the wide
R.,Friedman,S.D.,Gry,C.,Jenkins,E.B.,Morton,D.C.,Savage,B.D.,
varietyofprofilesshowninFig.7.Thisbeliesourabilitytoin- Sonnentrucker,P.,Vidal-Madjar,A.,Welty,D.E.,&York,D.G.2002,ApJ,
ferthegeneralpropertiesofthehostgasfromstudiesalongany 577,221
single sightline, no matter how comprehensive; studies of indi- Savage,B.D.,Drake,J.F.,Budich,W.,&Bohlin,R.C.1977,ApJ,216,291
Schlegel,D.J.,Finkbeiner,D.P.,&Davis,M.1998,ApJ,500,525
vidual absorption sightlines must be viewed demographically,
Sonnentrucker,P.,Welty,D.E.,Thorburn,J.A.,&York,D.G.2007,Astrophys.
asdatapointswithinalargefamilyofpossibleoutcomes,evenin J.,Suppl.Ser.,168,58
nominallysimilarconditions.Evenbeyondthis,therearesome Tachihara,K.,Abe,R.,Onishi,T.,Mizuno,A.,&Fukui,Y.2000,Publ.Astron.
obviousfundamentallimitstotheuseofabsorptionlinestode- Soc.Jpn.,52,1147
VanDishoeck,E.F.&Black,J.H.1986,Astrophys.J.,Suppl.Ser.,62,109
rivethepropertiesoftheinterveninggas.Inthepresentcaseonly
—.1988,ApJ,334,771
amapcouldcorrectthefalseimpressionthatthestarisocculted
Wannier,P.,Penprase,B.E.,&Andersson,B.-G.1997,ApJ,487,L165
bytwoseparateforegroundclouds.Likewise,theturbulentflows Wannier,P.G.,Penzias,A.A.,&Jenkins,E.B.1982,ApJ,254,100
in the foreground gas appear clearly in maps of the gas but not Wolfire,M.G.,Hollenbach,D.,McKee,C.F.,Tielens,A.G.G.M.,&Bakes,E.
atmostindividualpositions,therebyconveyingthefalseimpres- L.O.1995,ApJ,443,152
Wolfire,M.G.,McKee,C.F.,Hollenbach,D.,&Tielens,A.G.G.M.2003,ApJ,
sionofanoverly-quiescentmedium.
587,278
We intend to map with higher angular resolution some re- Wood,K.,Haffner,L.M.,Reynolds,R.J.,Mathis,J.S.,&Madsen,G.2005,
gionsoftheL121gaswhoselineprofilesarethermalatthe2.7(cid:48) ApJ,633,295
NANTENresolution,toseewhatkindofsubstructuremightbe
presentwhensomeformsofline-broadeningareabsent.Wewill
alsomapsomeL121gaswhoseprofilesarebroaderandwhose
velocitygradientsarenotfullyresolvedinFig.7at4(cid:48) (0.16pc)
beam-spacing.
This is the second paper in a loose series (see Pety et al.
(2008))whichwillalsoreportobservationsofsimilarkinemat-
ics on smaller angular scales and at higher angular resolution
6(cid:48)(cid:48)-22(cid:48)(cid:48)inotherdiffusecloudsofunknowndistancewhosepres-
encewasfirstmanifestedinourmm-waveabsorptionstudiesof
polyatomicmolecules.Inasubsequentpaperwewilldiscussthe
geometricalandphysicalinterpretationoftheinternalstructures
responsiblefortheflowsseenthereandinFig.7.
Acknowledgements. The National Radio Astronomy Observatory is operated
by Associated Universites, Inc. under a cooperative agreement with the US
NationalScienceFoundation.IRAMisoperatedbyCNRS(France),theMPG
(Germany)andtheIGN(Spain).ThisresearchmadeuseoftheSimbadastro-
nomicaldatabaseandtheNASAADSastrophysicaldatabasesystem.Thiswork
profitedfromdiscussionsofCOexcitationwithMichelGuelin.
References
Black,J.H.&Dalgarno,A.1977,Astrophys.J.,Suppl.Ser.,34,405
Burgh,E.B.,France,K.,&McCandliss,S.R.2007,ApJ,658,446
Cardelli,J.A.,Mathis,J.S.,Ebbets,D.C.,&Savage,B.D.1993,ApJ,402,L17
Crutcher,R.M.1979,ApJ,231,L151
Dobashi,K.,Uehara,H.,Kandori,R.,Sakurai,T.,Kaiden,M.,Umemoto,T.,&
Sato,F.2005,Publ.Astron.Soc.Jpn.,57,S1
Falgarone,E.,Panis,J.-F.,Heithausen,A.,Perault,M.,Stutzki,J.,Puget,J.-L.,
&Bensch,F.1998,A&A,331,669
Falgarone,E.&Phillips,T.G.1996,ApJ,472,191
Gaustad,J.E.,McCullough,P.R.,Rosing,W.,&VanBuren,D.2001,Publ.
Astron.Soc.Pac.,113,1326
Goldreich,P.&Kwan,J.1974,ApJ,189,441
Hartmann, D. & Burton, W. B. 1997, Atlas of galactic neutral hydrogen
(Cambridge;NewYork:CambridgeUniversityPress)
Herbig,G.H.1968,ZeitschriftfurAstrophysik,68,243
Knapp,G.R.&Jura,M.1976,ApJ,209,782
Kopp,M.,Gerin,M.,Roueff,E.,&LeBourlot,J.1996,A&A,305,558
Lambert,D.L.,Sheffer,Y.,Gilliland,R.L.,&Federman,S.R.1994,ApJ,420,
756
Langer,W.D.,Glassgold,A.E.,&Wilson,R.W.1987,ApJ,322,450
Liszt,H.,Lucas,R.,&Pety,J.2006,A&A,448,253
Liszt,H.S.1979,ApJ,233,L147