First observations of the water vapour line at 557 GHz from the interstellar medium

Abstract

. We report the first results of PIROG 7, a balloon-borne platform equipped with a 60 cm telescope and a heterodyne receiver designed for observations of the 557 GHz (1 10 ! 1 01 ) line of ortho-H 2 O. We describe the experiment, atmospheric transmission measurements, and astronomical observations of Orion. The lack of broad, centrally peaked line emission in this source, compared to the recently detected line of H 2 18 O at 548 GHz, indicates that the abundance of water is large and its emission optically thick. On the other hand, a weak and narrow absorption feature was detected toward Orion. While the signal-to-noise ratio of the observed absorption is such that it should only be considered as a tentative detection at the present moment, the result is consistent with the available information on the physical characteristics of the gas in this region. The feature can be interpreted as the result of absorption by a foreground ridge-like component of bright and wide emission arising f...

Full text

Page 1

thein terstellar medium
J?Tauber
?
?G?Olofsson
?
?G? Pilbratt
?
? L?Nordh
?
?andU?Frisk
???
?
ESAAstrophysicsDivision?SpaceScienceDepartment? ESTEC?P?O?Box????NL ?????AGNoordwijk?TheNetherlands
?
StockholmObservatory?S??????Saltsj?obaden?Sweden
?
Presentaddress?SwedishSpaceCorp oration?P ?O? Box ?????S? ?????Solna?Sw eden
Received??May ?????Accepted?? August????
Abstract?Wereportthe?rst results of PIR OG??a
balloon?borneplatform equipp edwitha ??cmtelescop e
anda heterodyne receiver designedforobservations ofthe
???GHz??
??
??
??
? lineofortho?H
?
O?Wedescrib e the
experimen t?atmospherictransmission measurements?and
astronomicalobservationsofOrion?The lack of broad?
cen trallypeaked lineemissioninthis source?compared
tothe recentlydetected lineofH
?
??
Oat???GHz? indi?
catesthattheabundanceofwateris largeandits emis?
sionopticallythic k?Ontheotherhand?aweak and nar?
rowabsorption featurewas detected towardOrion? While
thesignal?to?noise ratiooftheobserv edabsorption issuch
that itshould onlybeconsideredasatentativedetection
atthepresent momen t?the resultis consistent withthe
availableinformationonthephysicalcharacteristicsofthe
gasinthisregion?Thefeaturecanbeinterpretedasthe
resultofabsorptionbyaforeground ridge?like componen t
ofbrightandwideemissionarisingfromthewarmgasin
thehigh?velocitycoreofOrion?
Keywords?ISM?clouds?ISM?individual objects?Orion?
ISM?molecules? ISM? abundances?Techniques? spectro?
scopic?Telescop es
??Introducti on
Molecularlineemission isour mainlink tounderstanding
the physicalcharacteristics ofmolecularclouds?MCs??In
addition? tracemoleculessuch asCO?H
?
O?O
?
andothers
are themain coolan ts ofMCs ?Goldsmith? Langer??????
andma y th uspla ya major rolein theirdynamical ev olu?
tion? inparticularin thephasesleading to starformation?
H
?
Oemissionalmost certainlyhasan important rolein
lo calizedareasof MCsasdiv erseas photon?dominatedre?
gions?e?g?Tielens?Hollenbach??????shockedregions
Sendo?printrequeststo?J?T auber
?e?g?Draineetal???????and accretiondisks?Clearly?
determiningtheabundanceofthiskeycomp onentisa
paramoun tingredientinourkno wledgeofMCsandstar
formation?Consequently?observationsofwateremission
areoneofthemosteagerlyanticipated resultsofsubmm?
waveastronomy?
TheabundanceofH
?
OinMCs canbeinferred from
chemicalmodels ofgas?phaseprocesses?Thesemo dels
predictlargeabundancevariationswithtimeandphys?
icalconditions?Inaddition?thereareuncertaintiesresult?
ingfromourincompleteknowledgeofe?g?reaction rates
andbranchingratiosatlowtemperatures?grainsticking
anddesorption mechanisms? thein?uenceof shocks and
thecosmicra yionization rate?nottomention theinitial
c hemicalcomposition? Simplemodelsindicate?e?g?Mil?
lar?Nejad?????thatthegas?phaseabundanceofH
?
O
reachesapeakofX?H
?
O????
???
after? ??
?
???
?
years?long b eforec hemicalequilibrium hasbeen reached
?? ??
?
y ears?when theH
?
O abundanceis anorderof
magnitude low er??Inacollapsingcloud? whenthe dense
coreis accretinggasfrom thesurrounding protostellaren?
velop e?modelsindicate?e?g? Brownetal??????amuch
higherH
?
Oabundance?X?H
?
O?
?
?
??
??
?than inthe
coldsteady?state solution?Modelsof shockchemistry also
indicateanenhancementof H
?
Oabundancesbyoneto a
fewordersofmagnitude?e?g?Draineet al??????Mitchell
?????Gra??Dalgarno ??????ThewiderangeofH
?
O
abundancespredictedbythemodels issuchthat obser?
vationalconstraintsshouldbeusedtorestrictit?bothin
ordertoassessitsdynamicalimportance? andtousethis
speciesasaprobeofthephysical conditionsinthein ter?
stellarmedium?ISM??
Watervapourexistsintwostates?ortho? withthetwo
hydrogennucleispinsaligned?andpara?wherethey are
antialigned?Thetwostatesare notradiativelycoupled?
th usthetwoformsactastwodi?erentspecieswith nu?
clearspinstatisticalweightsof?and??Atthelowtemper?
?
X?H
?
O??
n?H
?
O?
n?H???n?H
?
?
?wheren isthen umber density

Page 2

observ
b
ations
observ
but
ed
a?ected
ard
y
star?forming
ties
regions?
inthe iso?
most
Itis
frequencies
also to estimate
Men
abundance
??????
via
of
ground?basedobservationschemicallyecies?
sible fromground?basedand airborneobserv atoriesb e?
cause ofabsorptionbypressure?broadenedlinesoftelluric
atmosphericwater? Onthe other hand?at thealtitudes
reac
serv
hable
ed
The
b
O
ystratospheric
in
observ
ballo
of
ons??
and
resp
??km??
estimate
ect
the tel?
ater
exci?
luric
abundances
tation
H
?
O lines are
the
O
narro
range
line
w enough
??
that
to
is
itb
in
ecomes
those
suc
feasible
regions
that
to attempt observations ofDoppler?shiftedlinesarisingin
the ISM?Such anexp erimentis described in this paper?At
airplanealtitudes?
?
?
??km?it isp ossibletoconduct obser?
v
serv
undertak
ations
ed
of
in
the
Orion
en
m uc
?T
carry
h
urner
rarerH
et
?
??
O
?????
ations
?
??
O?
Olofsson
of
??
O?????
?????
the
isotopic
Plam?
?
form?
b
The
?
?
??
?
??????
?
??
transition
ho
of ortho?H
HDO?H
?
?
??
O
ratio
emitsat
??? GHz? andhas recentlybeendetected ?Zmuidzinas et
al?????? inboth SgrB?andOrion?Abundanceestimates
of order
Finally
??
??
for
n
the
b
main
of
sp eciesare
w
inferred
ater
from
ha
these
already
are
w
b uncertain
topicratio andbeam?llingfactor?
possiblew
al?
ater
ofrelatedsp
mostnotablyH
?
O
?
?However?this isdi?cultb othob?
servationally?theabundance ofH
?
O
?
b eingestimated to
b
dance
eabout
of
three
H
orders
the
ofmagnitude
ISM?
lo
w
wer
er?
than
recen
for
t
w a?
ter
v
itself
ations
??
of
and
widespread
byuncertaintiesinthec hemical
?Cernic
mo
haro
delling?
et
Observations ofH
?
O
?
were?rst con?rmed byWoottenet
al????????Morerecently?Phillipsetal???????haveob?
H
?
?
anumb erclouds?
with
w
in
H
??
??
??
where theionisdetected?Anotherpossibilityis toinfer
H
?
Oabundances fromthose of HDO?Forexample?the
?
to
??
?
remedy
?
??
transition
this
ofHDO
the
at
pro
????
ject
GHz
describ
has
ed
b
here
eenob?
w
al?
observ
eckWrightwever?theOisun?
kno wn?and notb elievedtobe relatedtotheinterstellar
D?H???
??
ratioinasimple fashion?
?aumermaseringlinesve
eento several
oftenat?? GHz?cf?Reid?Moran??????butalsoat
higher ?e?g?tenet Because
thenatureoftheemissionprocess?andbecausemasers
arebelieved tobecon?nedto small?denseandhot sp ots?
theselinesprovide verylittleinformationonthe abun?
?
O ingeneralHo evobser?
maseremissional?
?????wouldputalowerlimitof??
??
on thegas?phase
waterabundancealongthemolecularridgeofOrion?
ationalevidencetolow
level
?
emissionpresentlyhonly
weak constrain tscanbeput onthegasphasewaterabun?
dance inthecold insterstellarmedium? As anattempt
situation?as
toout?
??
?
??
launc
ometer?
hed in
The
late
resulting
also
termediate
capable
frequency
of
?IF?
n
at
w
GHz
atvshorter
further
???
v
activelyformingstars?seee?g?theextensivereviewof this
regionbyGenzel?Stutzki??????Assuch?itmakesan
appropriateandin teresting?rstchoiceforatarget?The
Dopplershiftofemission linesin Orionwithrespecttothe
Earthissu?cientlylargethat itcanbeobservedduring
thespringlaunchwindow? In therestof thispaper?we
describethe?rstobservationsof Orion atthisfrequency?
theinstrumentused? and thelimitations imposedbythe
stillsigni?cantatmosphericopacity?
InadditiontothesearchforH
?
Odescribed here?two
satellitepro jectsare underwaydedicatedto observea
n
rms
umb
??
er ofimp
The
ortan
detector
tsubmillim
is
etre
hottky?dio
lines?including
de
the
er?
???
ated
GHz
at
w
om
ater line?SWAS ?the
and
Submillim
pump
eter?W
a
a
oscilla?
veAs?
tronomySatellite?isscheduled tobelaunched inlate?????
andOdin?a Swedish?In ternational submillimeterastron?
omyandaeronom ysatellite? in?????The Infrared Space
Observ
are
atoryof
in
the
the
Europ
mixer
ean
b
Space
a
Agency? dueto
terfer?
be
????? will
in
beobserving
signal
umerous
???
ater
is
lines
ampli?ed?
waelengths
and
than
do
?m?
erted
butwithapo orv elocity resolutioncomparedto hetero?
dyne observations?
??Theinstrumen t
The PIROG? experimentw as builtat theAstrophysics
Division oftheEurop ean SpaceAgency?andconsistedof
a??cmdiameterCassegraintelescope?followedbyahet?
erodynedetectionfront?endandanautocorrelationsp ec?
trometer?Thetelescop eisofanalt?az type?withane?ec?
tivef?d???andasurfaceaccuracyappropriateforopera?
tion in thesubmillimetre rangeofwavelengths?measured
?m??aScmixerop
rotemperature? edby lo cal
tor?LO? consistingofa phase?lockedGunndiodefollowed
byafrequencydoublerandtriplercombinationto achieve
therequiredfrequencyof??????GHz?TheLO andsignal
injectedtoyMartin?Puplettin
then wncon
to????MHzbeforebeingfedin tothespectrometer?
Thesp ectrometeris adigitalautocorrelatorofthehybrid
type?withamaximumbandwidth of ???MHz?splitinto
subbands of??MHz each? Duringthe experiment? itwas
con?guredformaximumbandwidthwithachannelreso?
lutionof???MHz?????kms
??
at???GHz??Thesystem
sensitivitywasmeasured inthelabto beT
Rx
?????K
DSBnoisetemperature?
Acommercialopticalscanningunitintegratedinto the
opticalb enchisusedtoshiftthesignalpathtooneof

Page 3

found toe
holm Observ?videsthe exp
withointrol?pandtelemetry?Attitude
iserformed erThenominalsta?
bilitachievwithsystem???rmsinb
imuth andelevation?oTV?camerasare attachedthe
sitions?thechopping p eriod isnormally keptat????ms?
Thus?choppingb etweenthetwocalibration loadsisused
to determinethe receiver sensitivitywhilethesystemsen?
sitivit
era
y?including
b
the
a
sky?
a
is determined
the
b
e
y
axis
chopping
in
b
t
e?
to
t
pro
ween
vide
ablank
con
p
uous
ositiononthe
trac
sky
king
and
during
theam
the
bien
submillim
ttemper?
etre
atureload? Normalastronomicalobserv ationsarecarried
outbychoppingbet weenthetwop ositionsonthesky?in
oneofwhic h isthesourceof in terest??Thecorrelator is
tied
axis
directly
w
to
measured
thescannerp osition
the
signal
?igh
and integrates
the
the
a
source
Gunn
andreference
and
signals
optical
independen
whic
tly
w
?
Theexperimentw as mounted on thePIROGgondola?
asecond generation ?ight?provenballo onplatform? con?
structedbytheSwedishSpaceCorporationfortheStock?
atory Thegondola proeriment
p tingconow er?con?
trolin azim uthis ac hievedbymeans ofagasjetservosys?
temwhichusesthesun asareference?Telescopeelevation
pbysteppmotors?pointing
yabletheis
?
othaz?
Tw to
platform asameans to achieve ?nep ointingaccuracy?one
ofthecameras is ?xedto the telescopeand alignedprevi?
ous tothe ?ighton astronomicalsources? Thesecondcam?
canepoin tedwy fromtelescop?igh
tinstar
observations?The relative alignment ofthe TV?cameras
was quan ti?edbeforethe?ightandchecked during?ight?
Therelativepointing ofthe camerastothesubmillim etre
as alsobeforetwithhelpof
oscillatorsourcehereplacedat
thefocusofanauxiliarylarge?apertureopticinthenear?
?eldofthetelescop e?Theuncertaintyin the pointingac?
curacyisdominatedbytheuncertainty inthekno wledge
ofthep osition of thesubmillimetreb eamrelative tothe
opticalaxis?whichis estimated at??
?
?
??Observ ations
The PIROG?exp erimentw aslaunched on?May????at
UT????fromtheCNESballoonrangelocatedatAire?
sur?l?Adourinsouthern France?Thelaunch timew as care?
fullychosenin ordertoobservethechosenastronomical
target?Orion?nearitsculmination?Launchandascent
were normalandceiling ?????km?wasreachedab out???
hoursafterlaunc h?Thegondolawas attitude?stabilizedfor
initialcalibrationsapproximately?hoursafterlaunch?In
total???? hours ofusefultime w ereobtained?ab outone
hourofwhichwassp entmeasuring theatmospheric emis?
sion? andthe rest ontheastronomical target?Orion?? The
therest frequencythewline?theof the IF
passband
ously The frequencyaccuracywbshift?
ingtheof theobservingcorre?
inAppendices?and?tothis paper?Theobservations of
Orionwerecarriedoutinthebeam?choppedmode?with
thesourceinitiallyinonebeam?andthenshiftedtothe
alternate
ditions
the
beam?
?oat
com
W
altitude?
bined
ealsocarried
A
the
outchopp
eginning
the
edobserv
of
to
ations
?igh
duce
ofaposition far?????
?
?fromOriontocon?rmthelack
oflineemissionin thereferenceb eam?Observ ationswere
ended atUT????whenthe pa yloadw asdroppedfrom
the balloon? Thelanding ofthegondolaw asvery smooth
and
the
the
impact
outgassing
core
damages
rate?
the
werelimited
er
line
to
temp
someexternal
eratures
opaque
com?
go
en
p
as
onen
T
ts?
Duringtheobservations?thepointingstabilit yw as
bwithinspeci?cations????
?
rms??theabsolute
pointinguncertaintyachiev edisestimatedat??
?
from
stellar calibrations? Thefrequencywasset toavaluenear
ofaterwidth
was suchthatboth thetelluriclineand the
Doppler?shifted Orion linewere presentin it simultane?
?ascon?rmedy
frequency LOandthe
sp ondingshift ofthe telluricwaterline inthe correlator
passband? Boththeinstrumentand system sensitivities
w eremeasuredfrequently?The receiverperformedb etter
than
the
inthe
near
lab
IRc?
oratory
??
?due
km
tothe
and
dry
the
atmospheric
orbital
con?
cit
att
date
the bthe
pro
t
a
hea vyoutgassingofwaterinsidethereceiver enclosure
wasobserv ed?whic hwasalleviatedby?ushingitwithN
?
gas?Howev er?theinstrumentsensitivitysteadily improved
throughout
MHz?
the?igh
the
t?probably
Orion
due
in
toaslow
In
decrease
fact?
in
Receivnoise
completely
as
ev
o
at
d
Rx
????? K ?DSB?w eremeasured nearthe endof
the?ight? whenthe b estOriondata w eregathered?Onthe
otherhand?imp ortantstanding?wavee?ectsconfused the
interpretationofthedata?asdescribedfurtheron?The
overallsensitivityofthesystem?includingthesky emis?
sion?wasmeasured atT
CW
????? K?DSB?neartheend
of the?igh t?
??Results
Atotalof ???hourswerespen tobservingthecoreof Orion?
ThesourcecoordinatesusedwerethoseofIRc??????
h
??
m
??
s
?????
?
??
?
??
??
?epoch ?????? TheLSRv elocity of
gas?s
??
?veloyof
Earthonof?ight
DopplershiftoftheexpectedOrionlineof??? MHzfrom
therestfrequency?At?igh taltitudeand??
?
elevation?
theFWHMofthetelluricbrightnessline pro?le is????
sothatlineisitswing?while
oftelluricis

Page 4

Fig? ??We show theaverageof allthedata observedinOrion?inaraw??P
source
?P
sky
??P
sky
?form?normalizedby theav erage
ofthe rawcalibration spectrum?Thev ertical linesmark theedgesandcentresofeachsubbandintheautocorrelator?each
subbandcovers??channelsor??MHz?Spuriousfeaturesmaybeexpectedat these locations?The black trianglesmarkthe
locationofsuchfeaturesasiden ti?edondata otherthan thoseav eraged here?Forreference?thev
LSR
orOrioncorrespondsto
channel ?????
?oataltitude?atthe frequency oftheOrion emissionthe
sky opacityis only ?????transmission??????
After discardingan umb er ofscans takenat theb e?
ginningof theobserv ations?which are clearly a?ectedby
thepresenceofwatervapourinsidethe enclosure?ab out
?? minutesworthofintegration?source?reference?re?
mainas ?good?data? Therawaverageofthesescans
isshowninFigure ??Strongbaselinee?ectsareclearly
present?consisting mainlyofa large?amplitudestanding
wave? Thewavelengthofthestandingwavesuggeststhat
itwascausedby re?ectionsfrom themylarwindowsepa?
ratingthefront?endenclosurefromthetelescope? In Fig?
ure?weha vemark edthelocationsof the edgesandcen?
tresofthe??MHz?wideautocorrelationsubbands?The
autocorrelatormaypro ducespuriousfeatures ordiscon ti?
n uitiesat these locations ?cf?Bos??????Wehavemarked
withblacktrianglesthelocationofsuchfeatures whic h
canbeidenti?edfromdatatakenduring the?ightbut
notwhileobservingthesource?i?e?whileintegratingo??
source?duringcalibrationscans?etc??Thev
LSR
ofOrion
??kms
??
?corresp ondstochannel?????thisis afa vourable
positionsinceitisinbetweentheedgeandcentreofasub?
band?andweexpectno spuriousfeaturesatthislocation?
For reference?thecentre ofthetelluricline is atc hannel
??????Notethata faintfeatureis visiblenear theOrion
channel? essentiallytheonly oneremainingsuperposed on
thestanding wavebaselineonce thekno wn autocorrelator
features areremov ed?
Thebaselinesho wninFigure? isextremelywell ?t
byasinglecosineand its ?rstharmonic? thus con?rming
itsstanding?waveorigin? Figure?showstheresult ofsuch
a?t?whic hhas?freeparameters?theamplitudesand
phasesof thecosineanditsharmonic?theperiod?anda
DCo?set??Thefaintfeaturedescribedabovenowappears
inthe?t residualsasarelativelynarrowabsorptionline
locatedb etweenchannels ??and ???If this ?line?were
arealsignalthen itshould notbeincludedinthe ?t to
thesinusoidal baseline? How ever? the ?tresults dep end
quitecritically onthen umb er ofc hannelsleft out ofthe
baseline?t?as isillustratedinFigure??There arereasons

Page 5

Fig? ?? Theav eraged rawdatais ?ttoasingle cosineandits?rst
??
harmonic?
Discussion
Wehave restrictedthechannelrangeslightlyto
avoidthestrongknownspuriousfeaturesnearchannelsNo?? and????seeFigure ???Theupp erpanelshowsthedataandthe
?ttedfunction?andthelower paneltheresiduals?Thetrianglesmarkthelocation oftheknown spuriouscorrelator features?
The?absorption?line featurelocatedbetweenchannels??and?? ismostprobablynotspurious? andis nearthelocationwhere
theemission fromOrionisexpected?To obtainT
A
?
atthefrequencyoftheabsorptionfeature?multiplytheverticalaxisunits
by???K?
toexpectthattheastronomical lineiswide?asdiscussedin
thenext section?andthusthis problemisnotatrivialone?
Thesafestapproachisto excludeonly anarrowfeature
fromthe?t?i?e?thetop tracein Figure???whichinfact
yieldsavirtuallyidenticalresulttothatinFigure??no
excludedc hannels??
Aftermultiplicationbyappropriate constants?the re?
sultingnoisermsinthebaselineportionoftheuppertrace
of Figure?is????K?inT
A
?
units?? stillalmosta fac?
toroftwohigherthan predictedfromthe systemtem?
peraturemeasuremen ts?Thermsin thebaseline ofthe
lowest traceofFigure?is only about??? higherthan
exp ected?This factmay indicatethatpartof thestruc?
ture seenintheresiduals corresp ondsto real signal?The
di?cultiesdiscussedabove donotw arrantexcessivebase?
line?tting? How ever?wehaveexperimentedwiththe?t
parameters ?gapwidthandlocation??andwe?ndthat
byadjustingthemappropriately? itispossibletoobtain
?emissionwings?oneithersideoftheabsorptionfeature?
Thesigni?canceofthisfact ismadeclearinthediscussion
?Section ???
Thesignal?to?noise ratiooftheabsorptionfeatureis
lo wenoughthatitsreality maybeputin toquestion?To
strengthenits case?weha ve splitthedataintoseveral
partsat random?andaveragedthe partsindependently?
The featureappearstopersist intheseaveragesand in?
dicatesthatit isnotan anomalyinarestrictedsubset
ofthedata?Inthenextsectionweshalldiscusswhether
theputativesignalisconsistentwithpredictionsforthis
source?
Thedetectionof the???GHzground?state transitionof
ortho?H
?
??
ObyZm uidzinasetal? ??????in Orion?KL
showsap eakbrigh tnessof????K in the???
?
beam of
the KAO? If theH
?
Oline at ???GHzwereoptically thin?
and ifweassume anabundance ratio
??
O?
??
Oequalto