DTIC ADA464679: ISO LWS Spectroscopy of M82: A Unified Evolutionary Model

10 Pages·0.19 MB·English

by  Defense Technical Information Center

THEASTROPHYSICALJOURNAL,511:721¨729,1999February1
(1999.TheAmericanAstronomicalSociety.Allrightsreserved.PrintedinU.S.A.
ISO LWS SPECTROSCOPY OF M82: A UNIFIED EVOLUTIONARY MODEL
JAMES W. COLBERT,1,2 MATTHEW A. MALKAN,1 PETER E. CLEGG,3 PIERRE COX,4 JACQUELINE FISCHER,5 STEVEN D. LORD,6
MICHAEL LUHMAN,5,7 SHOBITA SATYAPAL,8,9 HOWARD A. SMITH,10 LUIGI SPINOGLIO,11 GORDON STACEY,12 AND
SARAH J. UNGER3
Received1998May29;accepted1998September4
ABSTRACT
We present the —rst complete far-infrared spectrum (43¨197 km) of M82, the brightest infrared galaxy
in the sky, taken with the Long Wavelength Spectrometer of the Infrared Space Observatory (ISO). We
detected seven —ne structure emission lines, [O I] 63 and 145 km, [O III] 52 and 88 km , [N II] 122 km,
[N III] 57 km, and [C II] 158 km, and —tted their ratios to a combination starburst and photo-
dissociation region (PDR) model. The best —t is obtained with H II regions with n\250 cm~3, an ion-
ization parameter of 10~3.5, and PDRs with n\103.3 cm~3 and a far-ultraviolet (cid:209)ux of G \102.8. We
0
applied both continuous and instantaneous starburst models, with our best —t being a 3¨5 Myr old
instantaneous burst model with a 100 M cuto(cid:134). We also detected the ground-state rotational line of
_
OH in absorption at 119.4 km. No excited level OH transitions are apparent, indicating that the OH is
almost entirely in its ground state with a column density D4]1014 cm~2. The spectral energy distribu-
tion over the long-wavelength spectrometer wavelength range is well —tted with a 48 K dust temperature
and an optical depth, q P j~1.
Dust
Subject headings: dust, extinction ¨ galaxies: individual (M82) ¨ galaxies: ISM ¨
galaxies: starburst ¨ infrared: galaxies
1. INTRODUCTION the prototypical starburst galaxy. Sometime around 108 yr
ago, M82 experienced a close encounter with M81 (Yun,
Infrared luminous galaxies emit an infrared luminosity
Ho, & Lo 1993). This gravitational interaction likely pro-
comparabletoorgreaterthantheiropticalluminosity.The
ducedthebarseen(Telescoetal.1991),whichprovidedthe
starburst galaxy is an infrared luminous galaxy converting
large streaming motions necessary to funnel large amounts
its molecular interstellar medium (ISM) into stars at a rate
ofmoleculargastothegalaxyˇscenter.Thislargereservoir
that can not be sustained for a Hubble time. The radiation
ofmoleculargas,D2]108M (Wildetal.1992),provides
from these new stars is reprocessed into infrared radiation _
thefuelfortheongoingstarburst.
by the dust in their parental molecular clouds. Other pos-
StudiesoftheM82nucleusinthemid-infrared(Telescoet
sible energy sources for infrared luminous galaxies are
al.1991)sawstrongo(cid:134)-centerhotspots.Usingthevelocity
activegalacticnucleiortheenergyofshocksresultingfrom
pro—lesoftheM82far-infrared—nestructurelines,Lordet
galaxy interactions. Recent studies (e.g., Genzel et al. 1998)
al. (1996) modeled two hot spots as well. These hot spots
support the hypothesis that most of the brightest infrared
maybestarformationsitesresultingfromcloud-cloudcolli-
luminous galaxies are predominantly powered by recently
sionsthatoccurbecauseoforbitalcrowdingneartheinner
formed massive stars; however, the simultaneous presence
Lindblad resonances (Kenney et al. 1992). Alternatively,
of active galactic nuclei and active star formation in some
Satyapal et al. (1997) modeled the nucleus of M82 as a
galaxies shows that both processes can occur in the same
region of outward star propagation, with the hot spots
phaseoftheevolutionofluminousinfraredgalaxies.
beingthemostrecentandbrightestareasofstarformation.
As a result of its proximity (3.63 Mpc; Freedman et al.
Relatively insensitive to extinction, far-infrared spectros-
1994)andmoderateinfraredluminosity,M82isthebright-
copy can provide a unique probe of infrared-bright, dust-
est galaxy in the infrared. Because of its recent episode of
obscured galaxies like M82. Line ratios may be used to
starformation(e.g.,Riekeetal.1993),itisconsideredtobe
constrain physical parameters and tend to be less vulner-
able to calibration uncertainties as well. It is also in the
1DepartmentofPhysicsandAstronomy,UniversityofCalifornia,Los
far-infraredthatsomeofthemostimportantcoolinglinesof
Angeles,CA90095.
2Email:colbert=astro.ucla.edu. theISMofgalaxies,the[OI]63kmand[CII]158kmlines,
3PhysicsDepartment,QueenMaryandWest—eldCollege,University appear.AcloseexaminationofM82willprovideatemplate
ofLondon,MileEndRoad,LondonE14NS,England,UK. for future comparisons to possible starburst galaxies,
4InstitutdˇAstrophysiqueSpatiale,BA(cid:252)timent120,UniversitedeParis
includingthoseathighredshift.
XI,F-91405Orsay,France.
5Naval Research Laboratory, Remote Sensing Division, 4555 Over- Our observations and the analysis used on the M82 far-
lookAvenueSW,Washington,DC20375. infraredspectrumaredescribedin(cid:176)2,and(cid:176)3presentsour
6InfraredProcessingandAnalysisCenter,CaliforniaInstituteofTech- —ne structure line results. We discuss line ratio model
nology,Pasadena,CA91125.
—tting,OHabsorption,and—ttingoftheentirespectrumin
7NRC-NRLResearchAssociate.
(cid:176)4;(cid:176)5presentsour—nalconclusions.
8GoddardSpaceFlightCenter,Code685,Greenbelt,MD20771.
9NRC-NASAResearchAssociate. 2. OBSERVATIONS AND ANALYSIS
10Harvard-Smithsonian Center for Astrophysics, 60 Garden Street,
Cambridge,MA02138. We present Infrared Space Observatory (ISO) long-
11CNR-Istituto di Fisica dello Spazio Interplanetario, CP 27, 00044
wavelength spectrometer (LWS; Kessler et al. 1996; Clegg
Frascati,Italy.
et al. 1996) grating mode (43¨196.7 km, j/*jD200) obser-
12DepartmentofAstronomy,510SpaceScienceBuilding,CornellUni-
versity,Ithaca,NY14853. vations of M82. These observations were taken as part of
721
Report Documentation Page Form Approved
OMB No. 0704-0188
Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and
maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,
including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington
VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it

does not display a currently valid OMB control number.
1. REPORT DATE 3. DATES COVERED
1999 2. REPORT TYPE 00-00-1999 to 00-00-1999
4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER
ISO LWS Spectroscopy of M82: A Unified Evolutionary Model
5b. GRANT NUMBER
5c. PROGRAM ELEMENT NUMBER
6. AUTHOR(S) 5d. PROJECT NUMBER
5e. TASK NUMBER
5f. WORK UNIT NUMBER
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION
Naval Research Laboratory,4555 Overlook Avenue, REPORT NUMBER
SW,Washington,DC,20375
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)
11. SPONSOR/MONITOR’S REPORT
NUMBER(S)
12. DISTRIBUTION/AVAILABILITY STATEMENT
Approved for public release; distribution unlimited
13. SUPPLEMENTARY NOTES
14. ABSTRACT
15. SUBJECT TERMS
16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF 18. NUMBER 19a. NAME OF
ABSTRACT OF PAGES RESPONSIBLE PERSON

a. REPORT b. ABSTRACT c. THIS PAGE 9
unclassified unclassified unclassified
Standard Form 298 (Rev. 8-98)
Prescribed by ANSI Std Z39-18
722 COLBERT ET AL. Vol. 511
the Long Wavelength Spectrometer Consortiumˇs Extra- with a well-measured period of 3.54 cm~1. We used the
galactic Guaranteed-Time Program. All observations were sine-wave¨—ttingalgorithmforextendedsourcesinISAPto
madein1996Mayontwoseparateorbits.Thecentralposi- divideoutthefringes.Thedefringinghaslittlee(cid:134)ectonthe
tion was R.A. (2000)\9h55m52s.3 and decl. (2000)\ majority of line (cid:209)uxes, altering the weak 145 and 122 km
69¡[email protected]. Our main observations consisted of 20 full lines by only 5%. Only the OH absorption at 119 km
gratingscans,withatotalintegrationoverallscansof8sat changes signi—cantly, becoming 33% weaker. Finally, we
each wavelength position spaced at one-quarter of the used a 3 p median clip on the data before averaging it all
resolutionelement.Thespectralresolutionwas0.29kmfor intoasinglespectrumwithbinsof0.05km.
the 43¨93 km range and 0.6 km for the 80¨196 km range.
3. FINE STRUCTURE EMISSION LINES
Also,inthe—nalaveragedspectrum,thespectralresolution
between 80 and 93 km lies between these two values, Table 1 lists the seven —ne structure lines detected along
resultingfromouraveragingoftheoverlapoftwodetectors with 1 p uncertainties. The measured wavelength of each
that are detecting di(cid:134)erent orders of the grating. We also line is consistent with a 225 km s~1 redshift. The qˇs listed
tookaspectrumo(cid:134)setby*a\]0s.8 and*d\]17@from are optical depths at each wavelength assuming a dust
the center of the galaxy in order to measure the back- extinction model from Adams, Lada, & Shu (1988) and
ground. This o(cid:134)set observation had the same integration q Pj~1atwavelengthslongerthanD50km.Wecalibrate
j
timebutwasoveramorelimitedwavelengthrange.Alldata this model using the visual extinction from the study of
used in this paper were processed through the LWS Pipe- near-infrared hydrogen recombination lines by Satyapal et
lineVersion7. al. (1995). Assuming a foreground screen they found anA
V
The spectra were (cid:209)ux-calibrated with respect to Uranus thatrangedfrom2¨12mag,withitshighestvaluesnearthe
(Swinyard et al. 1998). The individual detector scans were nucleus where A D10. We adopt this central value of 10
V
calibrated to within ^4%¨6% of each other, based on forourvisualextinction,notingthatatthewavelengthswe
overlapping detectors, with the notable exception of detec- are considering small errors in A have little e(cid:134)ect on the
V
tor LW4, covering wavelengths 150¨170 km, about 15% —nal line (cid:209)uxes. The corrected (cid:209)ux is the measured (cid:209)ux
lower.Severalreliabledetectorswerechosen,andtherestof multiplied by exp(q), appropriate for an external dust
the detectors were multiplicatively shifted so that overlap- screen.Figure1displaystheentirespectrum.
ping points would match. The LWS beam is roughly 80A The—nestructurelineswepresenthavebeenmeasuredin
FWHM, but it does show some variation with detector, thepastbytheKuiperAirborneObservatory(KAO;Lord
rangingfrom65A¨85AFWHM(Swinyardetal.1998). et al. 1996; Petuchowski et al. 1994; Stacey et al. 1991;
We performed the postpipeline analysis with the ISO Du(cid:134)y et al. 1987; Lugten et al. 1986; Watson et al. 1984),
Spectral Analysis Package (ISAP). Even after pipeline pro- but most were taken on separate nights with di(cid:134)erent cali-
cessing the spectrum remains contaminated with bad data brators. This complicates line ratios, because beam sizes,
point(cid:147)(cid:147)glitchesˇˇfromcosmicrayhits,whichwecleanedby spectral resolutions, and (cid:209)ux calibration schemes di(cid:134)er.
plotting several scans on top of each other and looking for AnotheradvantageofthisspectrumoverKAOisthecom-
thecharacteristicfastrisingandslowlyfallingglitchshape. plete absence of telluric absorption features, which had
Allbaddatapointswerethenremovedaswellasanysuspi- greateste(cid:134)ectonthemeasurementsofthe[OI]63kmand
cious and widely deviating points following the glitch, [N II] 122 km lines. We have improved the signal-to-noise
wherethedetectorshowedmemorye(cid:134)ects.Roughly20%of ratio on all lines, but the greatest increases were for the
thedatawerediscardedasunusable. weak[NII]122,[NIII]57,and[OI]145kmlines.Table2
Raw data also contain a fringing pattern that is believed givestheKAOlinesandtheirmeasuredcontinuum.Several
toarisefrominterferencebetweenthebeamcomingo(cid:134)the of the KAO continuum measurements provided are
—eld mirror and another beam re(cid:209)ected by the substrate publishedhereforthe—rsttime.
holding that mirror. Fringing is only a minor e(cid:134)ect for the TheagreementbetweentheKAOmeasurementsandour
shorter wavelengths (\70 km): the fringing is weaker, and own (cid:209)uxes is satisfactory, with the most signi—cant di(cid:134)er-
the spectrum is inherently noisier. At wavelengths longer ences in the weak [N II] 122 and [O I] 145 km lines. Our
than100km,fringesapproachD5%ofthecontinuumand continuum measurements, however, go from being 30%
endanger our ability to measure accurately the continuum higher to 10% lower than the KAO values. Some of this
level as well as to identify possible weak emission and maybeduetothesmallerbeamsemployedforKAOwork
absorption lines. The fringe is sinusoidal in wavenumber (30A¨55A),butthatcannotexplainallthediscrepancies.
TABLE 1
EMISSION-LINEFLUXES
Restj Flux CorrectedFlux Continuum
Line (km) (Wcm~2) q (Wcm~2)a (Jy)b
[OIII]...... 51.81 10.3^0.5]10~18 0.10 11.3^0.6]10~18 1490
[NIII]...... 57.32 3.4^0.5]10~18 0.09 3.7^0.5]10~18 1695
[OI]....... 63.18 17.6^0.5]10~18 0.08 19.1^0.5]10~18 1860
[OIII]...... 88.36 8.6^0.4]10~18 0.06 9.1^0.4]10~18 1970
[NII] ...... 121.90 1.7^0.3]10~18 0.04 1.8^0.3]10~18 1460
[OI]....... 145.53 1.2^0.1]10~18 0.04 1.2^0.1]10~18 1155
[CII]....... 157.74 13.4^0.1]10~18 0.03 13.8^0.1]10~18 1005
aCorrected(cid:209)uxisthemeasured(cid:209)uxcorrectedforextinction.
bContinuum(cid:209)uxdensityhasnotbeencorrectedforextinction.
No. 2, 1999 ISO LWS SPECTROSCOPY OF M82 723
FIG. 2.¨Blackbody—t.Thesolidlineisactualdatacorrectedforpre-
dictedextinction;thedottedlineisabest-—tblackbodyspectrumwith(cid:209)ux
densityPB (T)(1-e~qDust)law,whereqPj~1.Forcomparison,thedashed
j
lineisourbest—twithqPj~2.
FIG. 1.¨LWSspectrumforM82.Seven—nestructurelinesarevisible,
aswellasanOHabsorptionat119km.Thestrongfeaturesseenaround
43¨50kmareunidenti—edandpossiblyartifactsofthe—rstdetector. If we then assume q \1 at D2 km (from A \10), we
Dust V
get a covering fraction of almost unity. Emissivity laws
where qPj~1.5 and j~2 were also considered. The j~1.5
The Infrared Astronomical Satellite (IRAS) measured a
law does not —t as well, and the j~2 law showed large
continuum(cid:209)uxof1271Jyat60kmand1351Jyat100km
deviations. The 48 K temperature is consistent with the
(Riceetal.1988),bothlowermeasurementsthanthosepre-
formula from Spinoglio et al. (1995), which determines the
sentedhere.SincetheIRASvaluesrepresentthetotalM82
color temperature of a galaxy with a j~1 emissivity law
(cid:209)ux in an area roughly 10@]5@ and since the background
fromitsspectralindexfoundusingonlythe(cid:209)uxdensitiesat
measured by IRAS and by us in reference positions is less
60 and 100 km. From our data a for M82 is 0.04.
than2Jy,thereasonforthisdiscrepancyisunknown. 60h100
When we substitute that into their regression formula,
Whilethepreviouslinework(Du(cid:134)yetal.1987;Lugtenet
T \11.4](a ]4.67) K, the result is 53 K, only
al.1986;Lordetal.1996)achievedcalibrationaccuraciesto color 60h100
5Kdi(cid:134)erentthana—ttotheentirefar-infraredspectrum.
the 20%¨30% level, our study provides the —rst complete,
Figure 2 shows the —t with extrapolation into the sub-
medium-resolution, far-infrared spectrum, with substan-
millimeter. Not shown is Hughes, Gear, & Robsonˇs (1994)
tially reduced uncertainty (D5%), no atmospheric prob-
datapointfor450 km,jF \(2.7^0.4)]10~17 Wcm~2,
lems, a standard aperture, and a single calibration scheme. j
measuring the (cid:209)ux from the central 68@@]68@@ of M82.
Furthermore, by using line ratios rather than absolute
Inputting450kmintoour—tgivesajF \5.5]10~17 W
(cid:209)uxes to determine starburst/photodissociation region j
cm~2,afactorof2high,indicating,perhaps,asteepeningin
(PDR) properties, most systematic uncertainties should
the j dependence for the dust emissivity law in the sub-
cancel.
millimeter.Klein,Wielebinski,&Morsi(1988),—ttingafew
4. MODELING AND OTHER RESULTS points from the infrared, submillimeter, and out into the
millimeter radio portion of the spectrum, did derive a
4.1. BlackbodyFitting steeperdustemissivitylaw,withq Pj~1.5.
Dust
We —t our spectrum with a function of the form F P The total (cid:209)ux for the infrared region observed is
j
B (T)(1[e~qDust), where we assume q Pj~1. This gave 8.15]10~15 W cm~2, which at a distance of 3.63 Mpc
j Dust
usagood—tforatemperatureof48Kfortheemittingdust. correspondsto3.2]1010L .Assumingitcontinuesshort-
_
TABLE 2
KAOCONTINUAANDFLUXES
KAOLinea KAOContinuum HPBW
Line (Wcm~2) (Jy) (arcsec) Reference
[OIII]52km ...... 9.5^0.7]10~18 1207 48 1
[NIII]57km ...... 3.9^0.4]10~18 1269 48 1
[OI]63km........ 14.2^3.4]10~18 1254 44 2
[OIII]88km ...... 8.6^0.5]10~18 1689 48 1
[NII]122km...... 2.9‘0.9]10~18 1190 45 3
[OI]145km ...... 0.84^~00.6.24]10~18 1130 55 4
[CII]158km...... 14^4.2]10~18 1150 55 4
aAllline(cid:209)uxeshavebeencorrectedforsameextinctionusedinTable1.
REFERENCES¨(1)Du(cid:134)yetal.1987;(2)Lordetal.1996;(3)Petuchowskietal.1994;(4)Lugten
etal.1986.
724 COLBERT ET AL. Vol. 511
ward of 43 km as a blackbody, the total far-infrared (cid:209)ux anddustgrainswereassumedfortheemittingHIIregions.
would be 9.7]10~15 W cm~2 or 3.8]1010L , in agree- The[OIII]52km/[OIII]88kmratioissensitivetodensity,
_
ment with the estimate by Telesco & Harper (1980). The but it is almost completely independent of ionization
total (cid:209)ux in the —ne structure lines is 6.0]10~17 W cm~2 parameter and model type. We found the best —t for a
or 0.6% of the infrared (cid:209)ux. The [C II]/F(FIR) ratio in the densityof250cm~3forallmodels.The[NIII]57km/[NII]
LWSbeamis1.4]10~3,similartothatseentowardgalac- 122 km ratio acts in an orthogonal manner to our —rst
tic H II regions and other starburst and normal galaxies diagnosticratio.Itisweaklydependentondensityforlow-
(Staceyetal.1991). densityHIIregions,butitisstronglydependentonioniza-
tionparameterandweaklydependentonmodeltype.Here
4.2. Emission-LineRatioModeling we—tlogUvaryingfrom[3.1forthe30M cuto(cid:134)contin-
_
uousstarformationto [3.5fora3¨7Myroldinstantane-
A realistic starburst of a given age and duration is inter-
ous burst with a 100 M cuto(cid:134). These —ts with di(cid:134)erent
mediatebetweenthesimplecasesofaninstantaneousburst _
ionization parameters demonstrate Spinoglio & Malkanˇs
andcontinuousstarformation.Inordertolimitthenumber
(1992) point that a harder ionizing spectrum can be some-
of free parameters and to bracket the real situation, we
whatcompensatedforbyamodestdecreaseinU.
lookedattwodi(cid:134)erentstarbursttypes,instantaneousburst
Changeininputabundancesmadelittledi(cid:134)erenceinour
andcontinuousstarformation.Foreachoneourcombined
derived starburst parameters. We ran CLOUDY models
starburst H II region and photodissociation region (PDR)
with Orion abundances, ISM abundances, solar abun-
model has six free parameters: density in H II region,
dances, and twice solar metallicity starburst abundances,
density in PDR, ionization parameter in H II region, far-
where the abundances are those de—ned in CLOUDY
ultraviolet(cid:209)uxinPDR,inputstarburstage,andinputstar-
(exceptfortheISMabundances,whichcomefromSembach
burstuppermasslimit.
& Savage 1996). No matter what abundances we input, we
Table 3 lists the measured line ratios of M82 together
always derived basically the same densities and ionization
with the predicted combination starburst (H II region) and
parameters. The quality of the —t did vary somewhat, but
PDR model ratios. Following the methodology of Fischer
this was mainly the —t of oxygen to nitrogen line ratios
et al. (1996), the H II region model was created using the ([O III] 52 km/[N III] 57 km, [O III] 88 km/[N II] 122 km,
programCLOUDYbyGaryFerland(Version90;Ferland
andsoon),whichonewouldexpecttochangewhenaltering
1996)andinputspectralenergydistributions(SEDs)ofstar-
the ratio of oxygen to nitrogen abundances. Orion abun-
bursts described in Leitherer & Heckman (1995). To test
dances —t best, matching to within 10% of our measured
di(cid:134)erentSEDmodels,wetriedtwouppercuto(cid:134)masses,30
ratios.ISMandsolarabundanceshaveonlyslightlydi(cid:134)er-
and100M ,andtwobursttypes,instantaneousburstsand
_ entO/Nabundanceratios,andsotheyalsoproducedsatis-
continuous star formation, for ages of 1¨25 Myr. All the
factory —ts (within 30%) of the measured O/N line ratios.
SEDmodelshadsolarmetallicityandinitialmassfunction
Twice solar metallicity starburst abundances have high
(IMF)slopesof2.35.Theinputparametersaredensityand
O/N abundance ratios and consequently produced high
theionizationparameter,U,whichisde—nedastheratioof
O/Nlineratios,D2.5]thatobserved.
ionizingphotonstohydrogenatomsattheinnerfaceofthe
OfthefourmodelinputstarburstSEDsconsidered,only
cloud:
two produced satisfactory ratio —ts: the 3¨7 Myr instanta-
Q(H) neous burst with 100 M cuto(cid:134) and the 8¨25 Myr contin-
U\ , (1) uous star formation wi_th 30 M cuto(cid:134). A —t was also
4nr2n c _
H achieved with the continuous star formation with 100M
_
where Q(H) is the number of (ionizing photons) s~1, n is cuto(cid:134),butonlyafterreducingthenitrogenabundancebya
H
thetotalhydrogendensityattheinnerfaceofthecloud,and third. We can further constrain the models by looking at
risthedistancefromtheionizingsource.Orionabundances their total bolometric luminosity. Using the extinction-
TABLE 3
LINERATIOS
LineRatio MeasuredRatio Instantaneousa Continuousb
[OIII]52km/[NIII]57km....... 3.0^0.4 3.0 3.1
[OIII]52km/[OIII]88km ...... 1.24^0.08 1.26 1.32
[OIII]52km/[OI]63km........ 0.59^0.04 0.59 0.59
[OIII]52km/[NII]122km...... 6.3^0.5 6.3 6.6
[OIII]52km/[OI]145km ...... 9.4^0.9 9.2 10.5
[OIII]52km/[CII]158km...... 0.82^0.04 0.82c 0.82c
[NIII]57km/[NII]122km...... 2.1^0.3 2.1 2.1
[OI]63km/[OI]145km........ 15.9^1.4 15.5 17.9
[OI]63km/[CII]158km ....... 1.38^0.03 1.39 1.41
[NII]122km/[NII]205km...... 4.1^1.4d 4.4 4.9
aInstantaneousburstmodel,3¨5Myroldwith100M cuto(cid:134)plusPDRcontribution,the
_
preferredmodel.
bContinuousstarformationmodel,8¨25Myroldwith30M cuto(cid:134)plusPDRcontribu-
_
tion.Thismodelwasrejected,becauseitproducedtoomuchluminosity.
cThisratioisforcedtomatchexactlybythe—ttingmethod.
dISO did not observe the [N II] 205 km line; we use the KAO observation from Petu-
chowskietal.(1994).
No. 2, 1999 ISO LWS SPECTROSCOPY OF M82 725
corrected Brc (cid:209)ux contained within an ISO LWS beam Kaufmanetal.(1998)modelsgiveFlux(CII158km]CII
(fromSatyapaletal.1997)andassuminganelectrondensity 63 km]O I 145 km)/FIR\3.9]10~3, roughly equal to
D200 cm~3 and an electron temperature D5000 K, we the2.9]10~3wegetaftersubtractingourHIIregionline
found an ionizing photon rate of 8]1053 s~1. This is contribution. It should be noted that the Kaufman et al.
similartovaluespreviouslyfound.Forinstance,McLeodat (1998) PDR models use Galactic ISM abundances, not the
al. (1993) found a rate of 1.05]1054 s~1 using a smaller Orion abundances input into our CLOUDY H II region
beam (30A) but di(cid:134)erent assumptions about extinction. As models. However, for our PDR density and G , the di(cid:134)er-
0
oneexperiment,wescaledtheLeitherer&Heckman(1995) enceinusingtwodi(cid:134)erentabundancesisnegligible.
models to this ionizing photon rate and found that the Sincetheregionswearemodelingareinphysicalcontact
continuous star formation produces D1011L , more than with one another, we should be able to make some contin-
_
twicethemeasuredfar-infraredluminosity(D4]1010L ). uity checks across the model boundaries. Following the
_
The older instantaneous bursts also produce too much treatment described by Satyapal et al. (1998), we start by
luminosity, but the more recent bursts, 3¨5 Myr old, only checkingthatthereispressureequilibriumbetweentheHII
produce (3¨5)]1010 L , more consistent with the region and the PDR. The edge of the H II region has a
_
observed total far-infrared luminosity. This makes the 3¨5 temperature of 4000 K, which gives a P/kD106 cm~3 K.
Myr instantaneous burst our preferred input SED. Small InputingourPDRsurfacetemperatureofD250K,derived
errorsinionizingphotonratedonota(cid:134)ectthisconclusion. from Kaufman et al. (1998), and density gives a P/kD
AsimilarcheckwasmadebycomparingtheLeitherer& 5]105cm~3KorroughlythesameasourHIIregion.The
Heckman (1995) starburst masses expected to the dynami- second boundary condition is that the far-ultraviolet (cid:209)ux
calmassderivedfromthemassmodelbyGotzetal.(1990), leavingtheHIIregionshouldequalthatenteringourPDR
which gives us 1.6]109 M within our ISO beam. model. We measure the far-ultraviolet (cid:209)ux leaving the H II
_
McLeod et al. (1993) also used the Gotz et al. (1990) mass regionbytakingtheCLOUDYoutputcontinuumandinte-
modelbutfortheirsmallerbeamsize.Aftertheysubtracted grating the (cid:209)ux from 6 to 13 eV. This depends strongly on
estimates for molecular mass and an older stellar popu- input SED, giving us G \102.2¨102.4 for the instantane-
0
lation, they arrived at an estimate of 2.5]108 M for the ousburstmodels.Consideringtheuncertainties,thisiscon-
_
starburstpopulation.Riekeetal.(1993)usedthismasslimit sistentwithourmodelG \102.8.
along with modeling to predict that a lower mass cuto(cid:134) Lesteretal.(1987)not0edthatthe[OIII]52km/[NIII]57
must exist. Our preferred instantaneous burst model pre- km ratio will equal the actual O‘‘/N‘‘ ratio within
dicts starburst masses of 0.5¨1.3]108 M , depending on ^50%,thankstothesimilarcriticaldensitiesforthermali-
_
lowermasscuto(cid:134)(1.0¨0.1M ), whichdoesnotcomeclose zation of their respective emitting levels and ionization
_
to exceeding the dynamical mass contained within the ISO potentials,whicharewithin25%ofeachother.Withknow-
beam.Evenifwetakethe2.5]108M usedbyRiekeetal. ledge of the electron density, which one can get accurately
(1993) as our total allowed starburs_t mass, we have no from the [O III] 52 km/[O III] 88 km ratio, a very precise
trouble—ttingthemassandseenorequirementforalower O‘‘/N‘‘ ratio can be found that holds true for a broad
masscuto(cid:134),althoughonemightpossiblyexist. range of electron temperatures. Following the calculations
While the CLOUDY models —t the starburst lines well, of Lester et al. (1987) and taking our density of 250 cm~3,
they produce little [O I] and [C II] line emission. The we—ndaO‘‘/N‘‘ratioof3.8.Thisisslightlyhigherthan
amount of [C II] 158 km line (cid:209)ux produced by the that reported by Du(cid:134)y et al. (1987) for M82, O‘‘/
CLOUDY models depends on the SED we input, but it is N‘‘\3.1.Bothratiosarehigherthantheaveragevalueof
D24%¨31%ofthetotalobservedline(cid:209)uxfortheinstanta- H II regions toward the Galactic center, where SO‘‘/
neous burst. Both [C II] and [O I] lines are strongly pro- N‘‘TD0.9¨1.4(Dinersteinetal.1984).Onepossiblecon-
duced in PDRs, which one would expect to —nd in the clusion that could be drawn is that the O/N ratio is truly
interfacebetweenHIIregionsandmolecularclouds.Weuse higher in M82. However, it is more likely that this is an
the PDR models of Kaufman et al. (1998). These models e(cid:134)ect of the local ionizing conditions. In low-ionization
have two free parameters: density and far-ultraviolet (cid:209)ux, conditions, i.e., cool stars, more nitrogen is doubly ionized
G , which is expressed in units of local Galactic far- than oxygen, because the ionization potential of N‘ is
0
ultraviolet (cid:209)ux, or 1.6]10~3 ergs s~1 cm~2. Once again smallerthanthatofO‘.Thedi(cid:134)erenceinO‘‘/N‘‘ratio
thelineratiosprovideexcellentdiagnosticsfordetermining in the Galaxy would result from less massive (cooler) stars
these parameters. The ratio of [O I] 63 km/[O I] 145 km creating the H II regions. This is further supported by our
acts orthogonally to the [O I] 63 km/[C II] 158 km ratio model —t, where we —t the O‘‘ and N‘‘ lines for M82
over the densities and far-ultraviolet (cid:209)uxes considered with a high-ionization starburst SED and Orion abun-
(Wol—re,Tielens,&Hollenbach1990).Ifweassumethatthe dances without any oxygen overabundance. In fact, as this
remaining69%¨76%ofthe[CII]notaccountedforbythe sectionmentionedearlier,themoreoxygen-richabundance
starburstHIIregionmodeliscomingfromthePDRs,then modelsproducedsigni—cantlyworse—tstoourlineratios.
we have our PDR [O I] 63 km/[C II] 158 km ratio of Examining our best —t we come within 10% of all the
D1.6¨1.8. Combining that with the observed [O I] 63 ratios and closer on most. The derived parameters¨H II
km/[O I] 145 km ratio (D16) gives us a density of 103.3 region density, PDR density, ionization parameter, far-
cm~3 and a G around 102.8. These combined H II region ultraviolet(cid:209)ux,burstuppermasscuto(cid:134),andburstage¨are
0
and PDR models —t not only the line ratios, but, scaled to well determined, with even small changes in any of these
the distance and size of M82, they also —t the lines and inputscreatingratiosoutsidetheuncertaintiesoftheratios
continuumseen. measured. The majority of error still present in these —ts
We can check our derived G and PDR density using comesfromthemodelsthemselvesandtheassumptionswe
0
anotherPDRdiagnostic,theratioofthePDRline(cid:209)uxesto made, not from the inaccuracy in our line ratio measure-
the infrared continuum (cid:209)ux. For our parameters the ments.
726 COLBERT ET AL. Vol. 511
Wecancomparethisworkwithpreviousmodels.Spino- scans of the [O I] 63 km line does not indicate any self-
glio&Malkan(1992)also—tfar-infraredlineswithacom- absorption. Any signi—cant [O I] self-absorption would
bination H II region and PDR model. While achieving alterthePDRmodelingbyincreasingthePDRdensity.
similar H II region densities, their M82 model had a much There is one more important forbidden line in the far-
largerionizationparameter(10~2.5).Thisisentirelydueto infrared, the [N II] 205 km line, which lies just beyond the
the hardness of the input SED. The Leitherer & Heckman LWSlimitof196.7microns.Petuchowskietal.(1994)mea-
(1995) SEDs are signi—cantly harder than the SED Spino- sured it using the KAO, and our model does predict its
glio&Malkan(1992)used.TheSpinoglio&Malkan(1992) relative (cid:209)ux. We include both on our line ratio table and
PDR parameters were also di(cid:134)erent (G \104, n\104h5 notetheymatchtowithintheuncertaintyoftheKAO(cid:209)ux
cm~3),butthiscanbemainlyattributed0tothenewerPDR measurements.The[NII]lineratioisthebestdensitytracer
modelsofKaufmanetal.(1998).TheLordetal.(1996)PDR forlow-density(n\500cm~3)HIIregions,sotheaccurate
parameters for M82(G \103, n\104 cm~3) come closer reproduction of this line ratio lends credence to our esti-
0
to our derived values, but they also di(cid:134)er because of the mate:n D250cm~3.
e
newermodelsused.
Satyapal et al. (1997) and Reike et al. (1993) previously 4.3. CrossScansoftheCIIandOILines
created models of the M82 starburst, both —nding typical
InadditiontothefullscanmadeatthecenterofM82,we
starburst ages of around 107 yr. Reike et al. (1993) com- collectedcross-scansofthe[CII]158and[OI]63kmlines
pared many observational parameters (M , mass, lumi-
K using the LWS02 Astronomical Observation Template
nosity, CO index, UV photons), —tting M82 with two
(AOT). The maps consist of two cuts across M82, one
separatestarbursts,the—rstbeing13¨30Myrago.Satyapal
acrossthemajoraxiswithaP.A.\55¡andtheotheracross
etal.(1997)lookedatindividualstarburstclusters,measur-
the minor axis, perpendicular to the —rst. Seven obser-
ing their CO indices and Brc equivalent widths. Modeling
vations were made in each cut, starting 150A from M82ˇs
thosefeaturesforeachindividualcluster,theyfoundarange
center and spaced 50A apart. Each line represents an inte-
inageof6]106yr,implyinganoutwardpropagatingstar-
grationof3s(spectralelement)~1.ThetwoplotsinFigure3
burst. For the sake of simplicity, we did not attempt to
show the relative strengths of the emission lines along the
combine multiple models with di(cid:134)erent starburst ages, but twoaxes.Onlyupperlimitswerefoundforthe[OI]63km
wenotemanyoftheindividualstarburstclusteragesfound
line at the points 150A from M82ˇs center and are therefore
bySatyapaletal.(1997)werearound5Myr.
not included on the map. The major axis cut runs from
One unknown that could throw o(cid:134) the models is the
southwest to northeast, and the minor axis cut runs from
degree of [O I] 63 km self-absorption in M82. Stacey et al.
northwesttosoutheast.
(1983) —rst observed the 145 km [O I] line from the PDR
Both lines weaken away from the center, but while the
associatedwiththeOrionAHIIregionanddiscoveredthat [C II] 158 km line fall-o(cid:134) is symmetrical about the minor
the [O I] 63, 145, and [C II] 158 km line intensity ratios axis,itisnotsymmetricalaboutthemajoraxis;the[OI]63
indicatedopticallythick(qD2)emissioninthe63kmline.
km line is not symmetrical around either axis. This asym-
Tielens&Hollenbach(1985)modelsalsoindicatedthatthis
metryisnotthatsurprising.Previousmaps,suchasthosein
should be so. Later, through more sophisticated modeling
Ha (Waller, Gurwell, & Tamura 1992), the near-infrared
involving the [13C II] line and more complete mapping in
(Satyapal et al. 1995), and the submillimeter (Hughes et al.
the[CII]and[OI]lines,Staceyetal.(1993)estimatedthat
1994),showanexcessin(cid:209)uxtothewest.Staceyetal.(1991)
the[OI]63kmhasanopticaldepthD3intheOrionPDR.
The [O I] 63 km line is seen in absorption in Arp 220
(Fischer et al. 1998) and in Galactic star formation regions
(Poglitsch et al. 1996; Baluteau et al. 1997). In addition,
Hermann et al. (1997) explained an unusually low [O I] 63
km/[O I] 145 km ratio in the Dark Lane region of the
Orion Molecular Cloud as a likely site of [O I] 63 km
self-absorption.
Onewouldexpect[OI]63kmabsorptiontooccurwhere
cold foreground oxygen lies between us and an emitting
source. However, the gas density must exceed the critical
density for collisional de-excitation (n D5]105 cm~3)
crit
in order for the absorbed photons to be removed from our
beam. At lower densities the photons are mainly scattered,
withlittlereductionintheobservedlineluminosity.Inour
Galaxy, such dense clouds rarely cover more than a small
fractionofthe[OI]¨emittingregion.SincetheISMinmost
starburstandnormalgalaxiesseemstobeconstructedof a
superposition of many such clouds, which usually do not
cover each other simultaneously in both velocity and area,
the small regions of [O I] self absorption should not be
important. A notable exception is Arp 220 (Fischer et al.
1998),whereitislikelytheFIRsourceislargelyenveloped
FIG. 3.¨Plotsofthestrengthsoftwolinesacrossthemajor(top)and
bycold,densecloudswithabundantneutraloxygen.There
minor(bottom)axisofM82.Thesolidlineisthe[CII]158kmline,andthe
isnoreasontothinkofM82assuchanextremecase,andin dashedlineisthe[OI]63kmline.TheLWSbeamisrepresentedbythe
fact our analysis of high signal-to-noise ratio Fabry-Perot dottedline.
No. 2, 1999 ISO LWS SPECTROSCOPY OF M82 727
fromionizedregionsfromMaddenetal.(1993),weget
hl A g /g
I (ionized)\ u l X EM ,
C‘ 4n n 1][(g /g)]1](n /n ) C‘
crit u l e crit
(2)
where A is the spontaneous emission coefficient for the
2P ¨2P transition(2.36]10~6s~1),g /g istheratioof
3@2 1@2 u l
statistical weights in the upper and lower levels (2), n is
crit
the critical density (D35 cm~3), X is the abundance of
C‘
C‘ relative to hydrogen (D3.3]10~4), and EM is the
emission measure. Assuming electron density is low com-
paredtothecriticaldensity,weonlyneedameasurementof
the emission measure to derive the expected (cid:209)ux from the
extended, di(cid:134)use ionized gas of M82. We derive the EM
from the 6 cm maps of Seaquist & Odegard (1991),
FIG. 4.¨Plotof[OI]/[CII]lineratioacrossM82.Thesolidlineisthe assuming the majority of the emission comes from thermal
ratioacrossthemajoraxis,andthedottedlineistheratioalongtheminor free-freeemission.Atroughly50Ao(cid:134)theminoraxiswe—nd
axis. anEMD1300cm~6pc,givingusanexpectedintensityfor
C‘of1.7]10~4ergss~1cm~2sr~1oranexpected(cid:209)uxin
our ISO LWS beam of 2.0]10~18 W cm~2. This number
is consistent with that observed, indicating that the major-
didsimilarcutsinthe[CII]158kmline,whichalsoshowed ity of the extended C‘ emission is coming from the di(cid:134)use
this slight asymmetry in the major axis. Lord et al. (1996) ionizedmedium.
diddetailedmodelingof[OI]63kmlinepro—le,attributing
4.4. OHabsorption
themajorityoftheline(cid:209)uxtotwoo(cid:134)-centerhotspots,with
Theonlycertainabsorptionfeatureobservedinthespec-
the western spot the brighter of the two. We examined the
trumofM82isduetothetwolambda-doublingrotational
far-infrared continuum in our cross scans and also found
transitions of OH at D119.4 km, which are unresolved at
more(cid:209)uxtothewest.
theLWSgrating-moderesolution.Thesearethetransitions
We measured the spectral index a between 60 and 100
km,whereF Pla, ateveryspotinourscans.Thisa from the OH ground state of 2n3@2 J\3/2 to the next
is largest 50Al to the west, where it is D0.4, not at6M0h8120ˇ0s highest energy state, 2n J\5/2, which have been seen
3@2
before in emission in the KL nebula (Storey, Watson, &
center,wherea isD0.1.Thishighera impliesa
60h100 60h100 Townes1981)andinabsorptiontowardtheGalacticcenter
higher average temperature, further indication of a strong
(Genzel et al. 1985). No other OH lines are observed in
hot spot to the west. In addition, we also looked at the
M82.Thereisadipatthelocationofthe53kmtransitions,
spectral index a between 100 and 175 km, where we were
but it is the same scale as the noise and must be treated as
interested in testing whether our single temperature black-
anupperlimit.Table4liststhe1pupperlimitsforthenext
bodywithemissivityPj~1would—tfurtheroutwardfrom
mostlikelyOHlines.
thegalaxyˇscenter.Thea atthehotspottothewest
100h175 Theequivalentwidthofanopticallythinlineis
is D1.6, which is consistent with our blackbody —t.
However, the a measured to the east and farther g j4
outward in the1g0a0lha1x7y5 is smaller than expected, possibly W \A i 0 N , (3)
ij g 8nc
indicatingagrowingcolddustcomponent. j
Mostinterestingisthevariationinthe[OI]63km/[CII] where A is the Einstein coefficient for spontaneous emis-
ij
158kmlineratio,visibleinFigure4.Assumingthetwolines sionbetweentwolevels,iandj;gistheirstatisticalweights;
comemostlyfromPDRs,tracingthe[OI]63km/[CII]158 j isthetransitionˇslinecenter;andNisthecolumndensity
0
km line ratios will tell us something about conditions in ofmaterial(Aller1960).TakingtheAcoefficientsfromDes-
M82onalargescale.Theratioshowsamaximumroughly tombesetal.(1977),onecancalculatecolumndensitiesand
50Ao(cid:134)centertothenorthwestandsouthwestofthegalaxyˇs columndensityupperlimitsforthelines.Threeofthetran-
center and shows a steady decrease outward. According to sitions, 53, 79, and 119 km, are all from the ground state,
the models of Kaufman et al. (1998), this ratio decreases but we only detect the 119 km line because the A-
with PDR density and far-ultraviolet —eld strength. The coefficients for 53 and 79 km give equivalent widths 20¨80
generaltrendofdecreasing[OI]63km/[CII]158kmratio times less than that of the 119 km transition, assuming the
with distance from M82ˇs center is then understandable.
TABLE 4
One would expect both density and far-ultraviolet —eld to
decrease with distance from the galaxyˇs center. The o(cid:134)- OHLINES
center maximum in the [O I] 63 km/[C II] 158 km ratio
Line EquivalentWidth A
mustindicatewhereconditionsofhighdensity,highG , or ij
0 (km) Transition (km) (s~1)
bothexist.
The [C II] 158 km line (cid:209)ux also appears more extended 53....... 2n J\3/2]n 2J\3/2 \0.003 0.04
3@2 1@2
thanthe[OI]63kmline(cid:209)ux,possiblyindicatingitsassoci- 79....... 2n J\3/2]n 2J\1/2 \0.004 0.033
3@2 1@2
ation with the —lamentary features observed in optical 84....... 2n3@2J\5/2]n3@22J\7/2 \0.004 0.5
119...... 2n J\3/2]n 2J\5/2 0.022^0.004 0.13
recombinationlines,aspreviouslysuggestedbyStaceyetal. 3@2 3@2
(1991). Taking the formula for [C II] intensity originating 163...... 2n1@2J\1/2]n3@22J\3/2 \0.005 0.054
728 COLBERT ET AL. Vol. 511
transitions are optically thin. This is consistent with the obstructing atmosphere. Therefore, the relative calibration
upperlimitsobserved.Thelackofthe84and163kmlines, oftheselinesissuperiortothatusedforprioranalysis.Also
despite having higher cross sections, demonstrates that the reported is the discovery of an absorption line of OH from
absorbingOHgasisalmostentirelyinitsgroundstate.The itsloweststate.TheupperlimitsforotherpossibleOHlines
column density of OH in its ground rotational state is are enough to determine that most of the absorbing OH is
D4.2]1014 cm~2. Further analysis will be presented in a in its ground state. The whole spectrum is well —tted by a
laterpaper. single temperature dust component with an optical depth
Besides the 119 km OH absorption lines and the seven q Pj~1,givingatotalinfrared(cid:209)uxof3.8]1010L .
Dust _
emissionlines,weidentifynootherlinesorotherfeaturesin We —tted the infrared line ratios of M82 with a six-
the spectrum. We should, however, point out some visible parameter combination H II region and PDR model. The
features in the spectrum that may or may not be real. The parameters for the best —tting model are H II region
large bumps at the shortest wavelengths, 43¨50 km, are density\250 cm~3, U\10~3.5, PDR density\103.3
presentinotherLWSspectra(Arp220;Fischeretal.1998). cm~3, G \102.8, upper mass cuto(cid:134)\100 M and
These features appear to be an artifact of that detector, age\3¨50Myr.The[CII]158kmlinecomesD25%_ from
SW1, but they are being investigated further. Two other the H II regions and D75% from PDRs. In our model, the
apparentfeaturesat56and112km,resultfromunexpected starburst mass is D(0.5¨1.3)]108 M , depending on the
_
HD lines in the Uranus calibration spectrum. The 56 km lowermasscuto(cid:134)chosen.Thisisasigni—cantpercentageof
lineisparticularlytroublesome,becauseitconfusesthesitu- the molecular mass (D2]108M , Wild et al. 1992), indi-
ationontheshortwavelengthsideofthe[NIII]57kmline, cating the M82 starburst will n_ot be able to continue
where there is a hint of an absorption, possibly a p-H O producing stars, through another instantaneous burst or
2
line.Harveyetal(1998)haveexaminedtheISO/LWSM82 otherwise,formuchlongerbeforerunningoutofmass.The
spectrum for evidence of atomic hydrogen recombination LWS cross-scan data support the model of strong, o(cid:134)-
lines in this range. None were detected. The upper limits, center hot spots seen in the mid-infrared (Telesco et al.
constrained by their very low line-to-continuum ratio, lie 1991) and modeled from velocity pro—les by Lord et al.
withintherangeofexpectedvaluessetbytheknowncenti- (1996). Our —t for a 3¨5 Myr instantaneous starburst is a
meter and decimeter lines and a spontaneous emission simpli—cation of the true situation, which is several regions
process.Otherpotentialfeaturesnotalreadymentionedare with di(cid:134)erent starburst ages. The single burst we modeled
absorptions at 65.1, 66.4, and 149.2 km. Further analysis should be dominated by the brightest and most recent
and possible identi—cation will come in a later paper. Any bursts,whicharepresumablythehotspotsthemselves.
undetected lines present in the M82 spectrum would have
to have (cid:209)uxes less than D5]10~19 W cm~2 at 50 km to Thisworkwasmadepossiblebythehardworkanddedi-
D2]10~19Wcm~2at180km. cationoftheISOteamandtheLWSconsortium.Wegrate-
5. CONCLUSIONS fully acknowledge the use of the Kaufman et al. (1998)
models. In particular we would like to thank Matt Green-
We have obtained an LWS full grating scan of the pro- house,MarkWol—re,andMichaelKaufmanfortheircom-
totypical starburst galaxy, M82. This spectrum reveals ments and help. This work was supported by the NASA
seven bright —ne structure lines, previously detected from ISO program, including NASA grants NAG 5-3309,
theKAObutheredetectedatthesametime,withthesame NAGW-1711, and NAGW-4038. J. F. and M. L. also
instrument, with roughly the same beam, and without any acknowledgesupportbytheOfficeofNavalResearch.
REFERENCES
Adams,F.C.,Lada,C.J.,&Shu,F.H.1988,ApJ,326,865 Leitherer,C.,&Heckman,T.M.1995,ApJS,96,9
Aller, L. H. 1960, Stellar Atmospheres, ed. Greenstein, J. L. (Chicago: Lester,D.F.,Dinerstein,H.L.,Werner,M.W.,Watson,D.M.,Genzel,R.,
Univ.ChicagoPress),156 &Storey,J.W.V.1987,ApJ,320,573
Baluteau,J.-P.,etal.1997,A&A,322,L33 Lord,S.D.,Hollenbach,D.J.,Haas,M.R.,Rubin,R.H.,Colgan,S.W.J.,
Clegg,P.E.,etal.1996,A&A,315,L38 &Erickson,E.F.1996,ApJ,465,703
Destombes,J.L.,Marliere,C.,Baudry,A.,&Brillet,J.1977,A&A,60,55 Lugten,J.B.,Watson,D.M.,Crawford,M.K.,&Genzel,R.1986,ApJ,
Dinerstein,H.L.,Lester,D.F.,Werner,M.W.,Watson,D.M.,Genzel,R., 311,L51
&Rubin,R.1984,inAirborneAstronomySymposiumProceedings,ed. Madden,S.C.,Geis,N.,Genzel,R.,Herrmann,F.,Jackson,J.,Poglitsch,
H.A.ThronsonandE.F.Erickson(NASACP:2353),266 A.,Stacey,G.J.,&Townes,C.H.1993,ApJ,407,579
Du(cid:134)y,P.B.,Erickson,E.F.,Haas,M.R.,&Houck,J.R.1987,ApJ,315,68 McLeod,K.K.,Rieke,G.H.,Rieke,M.J.,&Kelly,D.M.1993,ApJ,412,
Ferland,G.J.1996,Univ.KentuckyDept.Phys.Astron.Int.Rep. 111
Fischer,J.,etal.1998,inFirstISOWorkshoponAnalyticalSpectroscopy, Petuchowski,S.J.,Bennett,C.L.,Haas,M.R.,Erickson,E.F.,Lord,S.D.,
ed.A.M.Heras,K.Leech,N.R.Trams,&M.Perry(Noordwijk:ESA), Rubin,R.H.,Colgan,S.W.,&Hollenbach,D.J.1994,ApJ,427,L17
149 Poglitsch,A.,Herrmann,F.,Genzel,R.,Madden,S.C.,Nikola,T.,Tim-
¨¨¨.1996,A&A,315,L97 mermann,R.,Geis,N.,&Stacey,G.J.1996,ApJ,462,L43
Freedman,W.,etal.1994,ApJ,427,628 Rieke,G.H.,Loken,K.,Rieke,M.J.,&Tamblyn,P.1993,ApJ,412,99
Genzel,R.,etal.1998,ApJ,498,579 Rice, W., Lonsdale, C. J., Soifer, B. T., Neugebauer, G., Kopan, E. L.,
Genzel,R.,Watson,D.M.,Crawford,M.K.,&Townes,C.H.1985,ApJ, Lloyd,L.A.,DeJong,T.,&Habing,H.J.1988,ApJS,68,91
297,766 Satyapal, S., Watson, D. M., Pipher, J. L, Forrest, W. J., Greenhouse,
Gotz,M.,McKeith,C.D.,Downes,D.,&Greve,A.1990,A&A,240,52 M.A.,Smith,H.A.,Fischer,J.,&Woodward,C.E.1997,ApJ,483,148
Harvey, V. I., Satyapal, S., Smith, H. A., and Strelnitski, V. 1998, AJ, Satyapal,S.,etal.1995,ApJ,448,611
submitted Satyapal,S.,etal.1998,inpreparation
Hermann, F., Madden, S. C., Nikola, T., Poglitsch, A., Timmerman, R., Seaquist,E.R.,&Odegard,N.1991,ApJ,369,320
Geis,N.,Townes,C.H.,&Stacey,G.J.1997,ApJ,481,343 Sembach,K.R.,&Savage,B.D.1996,ApJ,457,211
Hughes,D.H.,Gear,W.K.,&Robson,E.I.1994,MNRAS,270,641 Spinoglio,L.,&Malkan,M.A.1992,ApJ,399,504
Kaufman,M.,Wol—re,M.,Hollenbach,D.,&Luhman,M.1998,inprep- Spinoglio,L.,Malkan,M.A.,Rush,B.,Carrasco,L.,&Recillas-Cruz,E.
aration 1995,ApJ,453,616
Kenney,J.D.P.,Wilson,C.D.,Scoville,N.Z.,Devereux,N.A.,&Young, Stacey,G.J.,Geis,N.,Genzel,R.,Lugten,J.B.,Poglitsch,A.,Sternberg,A.,
J.S.1992,ApJ,395,L79 &Townes,C.H.1991,ApJ,373,42
Kessler,M.F.,etal.1996,A&A,315,L27 Stacey,G.J.,Ja(cid:134)e,D.T.,Geis,N.,Genzel,R.,Harris,A.I.,Poglitsch,A.,
Klein,U.,Wielebinski,R.,&Morsi,H.W.1988,A&A,190,41 Stutzki,J.,&Townes,C.H.1993,ApJ,404,219
No. 2, 1999 ISO LWS SPECTROSCOPY OF M82 729
Stacey,G.J.,Smyers,S.D.,Kurtz,N.T.,&Harwit,M.1983,ApJ,265,7 Waller,W.H.,Gurwell,M.,&Tamura,M.1992,AJ,104,63
Storey,J.W.V.,Watson,D.M.,&Townes,C.H.1981,ApJ,244,L27 Watson, D. M., Genzel, R., Townes, C. H., Werner, M. W., & Storey,
Swinyard,B.M.,etal.1998,inProc.SPIE,3354,inpress J.W.V.1984,ApJ,279,L1
Telesco,C.M.,Campins,H.,Joy,M.,Dietz,K.,&Decher,R.1991,ApJ, Wild, W., Harris, A. I., Eckart, A., Genzel, R., Graf, U. U., Jackson, J.,
369,135 Russell,A.P.G.,&Stutzki,J.1992,A&A,265,447
Telesco,C.M.,&Harper,D.A.1980,ApJ,235,392 Wol—re,M.G.,Tielens,A.G.G.M.,&Hollenbach,D.1990,ApJ,358,116
Tielens,A.G.G.M.,&Hollenbach,D.1985,ApJ,291,722 Yun,M.S.,Ho,P.T.,&Lo,K.Y.1993,ApJ,411,L17