PreprinttypesetusingLATEX styleemulateapjv. 9/08/03

ENVIRONMENTAL EFFECTSON GASEOUSDISKS OFTHE VIRGO SPIRAL GALAXIES1

HIROYUKI NAKANISHI1,2, NARIO KUNO2,3, YOSHIAKI SOFUE1, NAOKO SATO4,5, NAOMASA NAKAI6, YASUHIRO SHIOYA7,8,TOMOKA TOSAKI9,2, SACHIKO ONODERA1, KAZUO SORAI4, FUMI EGUSA1, AND AKIHIKO HIROTA10,2

ABSTRACTWe found high molecularfractions ( fmol; ratio of themolecularto total gassurfacedensities)in threeof fiveVirgospiralgalaxies in spiteof their low totalgascolumndensity, basedon 12CO

�J � 1 � 0� observationswith

theNobeyama45 m telescopeequippedwith a multi-beamreceiver, BEARS.We interpret this asa resultofenvironmental effects.Combining theCOdatawith HI data,therelationshipbetweenthesurfacedensityof thetotal gas(HI plusH2) and fmol indicatesthat thethreegalaxiesneartheclustercenterhave larger f mol valuesthanexpected for field galaxies,while theothersshow normal f mol. Thelarge fmol is interpretedasbeingdueeitherto effective HI gasstripping, evenin theinnerdisks,or to large ISM pressure inducedby thehigh ICMpressureand/orrampressure,although thepossibilityof anunusuallyhighmetallicitycannotberuledout.Subject headings: galaxies:clusters:individual (Virgo), galaxies:ISM, ISM: molecules,radiolines: ISM

1. INTRODUCTION

Galaxymembersin a clusterof galaxiesareembeddedin ahot diffuseintra-clustermedium(ICM) emittingX-rays,andsuffering variousenvironmentaleffects.SinceVirgois oneofthe nearest clusters,we canstudy the environmentaleffectstherein greatestdetail.

Observation in the 21 cm line show that the HI gasis de-ficient for galaxiesnearthecenterof theVirgo Cluster(e.g.,Cayatteet al. 1990). This is interpretedasbeinga resultofram pressurestripping. Vollmer et al. (2001) simulatedtheram pressureeffect by a numerical calculationandmatchedwell theobserveddistortionof theHI diskdueto therampres-sure.Kenney & Young (1988, 1989) observed theVirgo spi-ral galaxies in theCO

�J � 1–0) line usingthe14 m telescope

of theFive CollegeRadioAstronomy Observatory (FCRAO)andfound that the molecular gascontent appears to be nor-mal, even in HI deficientgalaxies. They reported that HI-deficientgalaxieshaveahighratioof CO flux to HI flux.

Whethera galaxysuffersenvironmental effectsor not, theinnerdisk of a spiral galaxyis dominated by molecular gas,becausein general the innerdisk is originally dominatedbymolecular gas.Therefore,it hasnotbeenclarifiedwhetherthe

Electronic address:hnakanis@nro.nao.ac.jp1 Instituteof Astronomy, TheUniversityof Tokyo, 2-21-1Osawa,Mitaka,

Tokyo 181-0015,Japan.2 NobeyamaRadio Observatory, Minamimaki, Minamisaku,Nagano384-

1305,Japan.3 The Graduate University for Advanced Studies(SOKENDAI), 2-21-1

Osawa,Mitaka,Tokyo 181-0015, Japan.4 Divisionof Physics,Graduate Schoolof Science,HokkaidoUniversity,

Sapporo060-0810,Japan.5 Center for EducationandResearchof LifelongLearning,Wakayamauni-

versity, Wakayama640-8510,Japan.6 Institute of Physics,University of Tsukuba,Ten-nodai, 1-1-1 Tsukuba,

Ibaraki305-8577,Japan.7 Astronomical Institute,Graduate Schoolof Science,TohokuUniversity,

Aramaki,Aoba,Sendai980-8578,Japan.7 Department of Physics,Faculty of Science, Ehime University, Mat-

suyama790-8577,Japan9 Gunma Astronomical Observatory, 6860-86 Nakayama, Takayama,

Agatsuma,Gunma377-0702, Japan.10 Departmentof Astronomy, Schoolof Science,TheUniversity of Tokyo,

Bunkyo-ku, Tokyo 113-0033,Japan.1 A part of this work wascarried out underthe commonuseobservation

programat theNobeyamaRadioObservatory (NRO). NRO is abranch of theNational Astronomical Observatory.

innerdisk is affectedby environmental effectsor not.Wecanenumeratethreepossibilitiesfor thephysicalcondi-

tion of the innerdisk: (1) rampressurestrippingoccurs onlyin theouterHI disk,andtheinnerdisksuffersnoenvironmen-tal effect, (2) rampressurestrippingoccursin theinnerdiskaswell astheouterdisk,but only HI gasis selectively stripped,(3) theISM pressureof the innerdisk increasesdueto exter-nal pressures,andmolecular gasformation is enhanced. Themolecular gascan be dominant in the inner disk in any ofthesethreepossibilities.

In order to develop our understandingof thephysicalcon-dition of the gasin innerdisksof clusterspiral galaxies, wemust pay attentionto the molecularfraction ( f mol; ratio ofmolecular to total gas surfacedensity).In thefirst possibility,therewould beno differencein themolecular fraction, f mol,in theinnerdiskbetweenclusterandfield galaxies.However,the fmol of clustergalaxieswould be larger thanthatof fieldgalaxiesin thesecondandthird possibilities.

In this paper, we show resultsof the highest resolutionsingle-dish observations in the CO line of five Virgo Clus-ter galaxiesachieved with the Nobeyama 45 m telescopeequippedwith a multi-beamreceiver, BEARS.We alsocom-parethe CO datawith HI datawith almost the samereso-lution obtained with the Very Large Array (VLA) C andD-configurations. Thesedatasetsenable us to investigate themolecular fraction at eachpoint of the spiral galaxieswitha fine scale,while earlier researchhasdealt with the totalamountsof HI andCO(H2) gasesin galaxies.Basedoncluesfrom fmol, wequantitatively discussthephysicalcondition ofthegasdisksin termsof (1)HI stripping,(2)external pressuredueto the ICM or rampressure, (3) intrinsic metallicity in agalaxy, and(4) theUV field of a galaxy. Thedistanceof theVirgo clusteris takento be16.1Mpc (Fereareseet al. 1996).

2. OBSERVATIONSAND DATA

The12CO�J � 1 � 0� observationsof theVirgo Clusterspi-

ralsweremadefrom 2002 December to 2004 April with theNobeyama45m radiotelescope.Thehalf-powerbeam-widthwas15� � (1.2kpcat theVirgoCluster)at115GHz. Weusedafocal planearrayreceiver, BEARS(SIS25-BEamArray Re-ceiver System),whichconsisted-of 25beams(5 � 5) andwasoperatedin thedouble-side-band (DSB) mode(Sunadaet al.2000). Theseparation betweenthebeamswas41� � � 1. Twenty-fivedigital auto-correlatorswith a500kHz resolution (1.3 km

2 Nakanishi et al.

s� 1 at115GHz)anda512MHz coverage(1332 km s � 1) (So-rai et al. 2000a) wereusedasspectrometers.

Calibrationof the line intensitywasmadewith anabsorb-ing chopperwheelin front of thereceiver, which yielded theantennatemperatureT �A correctedfor theatmosphericandan-tennaohmic losses. Sincethe scaleof the intensity varieswith thebeams,mainlydueto thedifferenceof thesidebandratio betweenthe upper sideband(USB) andthe lower sideband(LSB),wescaledtheobservedDSBintensityto theSSB(singlesideband)intensityusingscalingfactors thatwerede-termined by comparing theCOintensitiesof NGC7538 mea-suredwith BEARSanda single-beamSSBreceiver, SIS100.The main beamefficiency of the telescopewasηMB � 0 � 40at 115GHz, andthe intensitygiven in this paper is themainbeambrightnesstemperature,definedby TMB T �A ηMB.

WemappedfiveVirgo galaxieswith agrid spacingof 10� � � 3.Thenumber of totalobserved pointsfor eachgalaxy was576andthemappedareawas3 � � 95 � 3 � � 95, exceptfor NGC 4254andNGC 4569, whosenumberof observedpointswas1056andthe areawas3 � � 95 � 7 � � 38. The observationsweremadein theposition-switching mode with anoff-positionat anoff-set of 7� from the centersof the galaxies in azimuth. Thetypical rms noiseof the main beambrightnesstemperaturewas∆TMB � 0 � 09 � 0 � 18 K pervelocity channel. Pointingoftheantennawascalibratedby observing thecontinuum pointsource3C273 at43GHz,andits typicalerrorwaslessthan5 � �(peakvalue). Our observedsampleswereselectedout of thegalaxies in Sofueet al. (2003), andtheobserving parametersarelistedin Table1.

— Table1 —

To compare with the distribution of CO, we usedthe HIdataof the target galaxies mapped with the VLA. The HIdatawereadoptedfrom Phookun,Vogel,& Mundy (1993) forNGC 4254, from Cayatteet al. (1990) for NGC 4402, NGC4569, andNGC 4579, andfrom Phookun & Mundy (1995)for NGC 4654. The spatialresolutionof the HI mapswasabout 20� � , which is closeto that of our CO maps. Table2summarizestheparameters of theHI data.

— Table2 —

Figure1 shows thedistribution of theCO integratedinten-sity, ICO �� TMBdv (K km s� 1), (color) andtheHI intensity(contours) of thefive galaxies. TheCO emissionsdistributein theinner disks,andtheextensionsaretypically up to radiiof 1 � � 5 (7 kpc).

TheHI gasin NGC4254andNGC4654is abundantanditsdistribution is strongly distorted.TheHI disksof NGC 4402andNGC 4569aretruncatedat theedgesof theCO disks.InNGC 4579, theHI gassurroundsthe centralCO disk wheretheHI gasis deficient.

— Figure1 —

3. ENVIRONMENTAL EFFECTSON GASEOUSDISKS

3.1. Molecular Fraction as a Function of the Surface Densityof the Total Gas

Thefractionof moleculargasto the total gas(HI gasplusH2 gas),fmol, is determined by theISM (interstellarmedium)pressureP, metallicity Z, and UV radiation U (Elmegreen

1993). Herethemolecular fraction f mol is definedas

fmol ΣH2

ΣHI � ΣH2 � (1)

with ΣHI and ΣH2 being the surface massdensitiesof HIand H2, respectively. Since the ISM pressureis approxi-matelyproportional to the square of the gassurfacedensity(Elmegreen1993),

PP0� Σ2

Σ20� �

ΣHI � ΣH2 � 2�ΣHI � ΣH2 � 20 � (2)

fmol is expressedasa function of Σ, whenZ andU aregiven.Honma, Sofue, & Arimoto (1995), Kuno et al. (1995), andSoraiet al. (2000b) investigatedfmol in nearby spiral galax-ies, and showed that it is well reproducedby a model ofElmegreen(1993). Figure2 shows model curvesof the Σ–fmol relationcalculatedbasedonElmegreen(1993), wherewescaleP, U , andZ with thevaluesof thesolarneighborhood:P0, U0, andZ0. The gassurfacedensityat the solarneigh-borhood, Σ0, is takento be8 M � pc� 2 (Sanders,Solomon, &

Scoville 1984), which is adoptedto be proportional to P 1� 20 .

We show model curvestakingUV radiations of 0 � 1U0, 1U0,and10U0 andmetallicitiesof 0 � 1Z0, 1Z0, and10Z0.

— Figure2 —

Figure 3 shows fmol againstΣ for eachobserved point ofthefive galaxies,wheretheCO-to-H2 conversionfactorwasadoptedto beXCO � 1 � 0 � 1020 H2 cm� 2 K � 1 km � 1 s (Nakai& Kuno 1995). The dashedcurves in all panelsdenote theaveraged fmol of NGC 4254 andNGC 4654. Themolecularfractions, fmol, of NGC 4254 andNGC 4654show similarcurves to the models with U � 1U0 andZ � 1Z0, which isexpectedfor field galaxies.

On thecontrary, theotherthreegalaxies(NGC 4402, NGC4569, andNGC 4579) presentlargermolecular fractions thanthedashedcurve. NGC 4402shows anextraordinarily largefmol in spiteof thesmallsurfacedensityof thegas.For NGC4569 theplottedpoints arescatteredon theuppersideof thedashedcurve. NGC 4579shows a very large f mol in spiteofthesmallsurfacedensityof thegasasNGC4402. In thethreegalaxies, fmol is alwayslarger thanthedashedcurve.

— Figure3 —

3.2. Origin of the Unusual Molecular Fraction in the VirgoSpirals

3.2.1. Ram pressure Stripping

Ram pressure stripping is a possibleinterpretation of theunusually large fmol of NGC 4402, NGC 4569, and NGC4579.

Rampressurestripping wouldoccurif thecriterion�ρICM

ρISM � � Rd � � δv

Vrot � 2 �1 (3)

is satisfied,whereρISM andρICM arethevolumenumberden-sity of theISM andtheICM, respectively, R is thegalactocen-tric radiusof theelement,d is thethicknessof thegasdisk,δvis therelative velocity of anISM cloudagainst theICM, andVrot is therotationalvelocityof thegalaxy.

For a simplecase,we considera gasdisk with thicknessd 0 � 05 kpc, rotatingat Vrot 200 km s� 1. For a typical

Environmental Effects onVirgoSpirals 3

ICM wind with ρICM 10� 4 H cm� 1, andδv 1000km s� 1

in theVirgoCluster(Hidaka& Sofue2002),weobtainfor thecriterionof whererampressurestripping canoccur,

ρISM � 0 � 05

�R

kpc ���H cm� 3 � � (4)

The HI gascanbe dominant in the inter-arm region evenin innerdisks,andits densitytypically rangesfrom 0 � 01 to 1H cm� 3. Hence,the HI gascanbe strippedeven in the in-nerdisks.On theotherhand, themolecular cloudcanhardlybestrippedbecausethegasdensityin themolecular cloudis10 � 1000 H2 cm� 3. Therefore, the HI gascanbe strippedselectively from a disk, although the molecular gasstill re-mainsin the disk. In this case,fmol in the Σ— fmol diagrammovestowardtheupper-left sideof theΣ— fmol curveof fieldgalaxies.

NGC 4402, NGC 4569, andNGC 4579arelocatedwithintheprojectedradiusof 2 � from M87 at thecenterof theclus-ter, andareknown to beHI deficient.Hence,they musthaveexperiencedrampressure stripping of theHI gas. We empha-sizethatin thesethreegalaxies,(1) theHI disksarerestrictedwithin thecentralCO disksof 1 � � 5 radii (7 kpc) (figure1), (2)the HI clouds cansatisfythe criterionof ram pressure strip-ping while themolecular clouds cannot, and(3) the Σ— f moldiagramsshow unusuallylarge fmol, indicating that the in-nerdisksarehighly HI deficient(figure3). Thesethreefactssupport thescenariothattherampressurestrippingof HI gasoccurredevenin theinnerdisks.

On the otherhand, NGC 4254andNGC 4654 arelocatedasfar as 3� from thecenterof thecluster. They show theHI distribution extending to the outerdisksandthe Σ— f molrelationbeingthesameasthatof field galaxies.Hence,mostof theHI gasmusthave not yet beenstrippedin spiteof thestrongly distortedHI disks.

3.2.2. Higher External Pressure due to the ICM-Pressure orRam Pressure

Another possibility for making fmol large is a higherex-ternalpressure dueto theICM-pressure,or therampressure,whichmightmaketheISM pressurelarger thanthatestimatedby thesurfacedensity.

The ICM pressureaffectsall gasdisks isotropically. Thetypicaldensityandtemperatureof theICM at theVirgoClus-ter areρ 10� 4 cm� 3 andT 107 K, respectively (Nulsen& Bohringer 1995). On the otherhand, the typical densityandtemperatureof theinterstellarmoleculargasareρ 102

cm� 3 and T 10 K, respectively. Sincethe pressureP isproportional to ρT , the ICM pressureandISM pressure arecomparableto eachotherneartheclustercenter.

The ram pressurewould alsoincreasethe ISM pressureiftheISM werenotstripped,asmentionedby Kenney & Young(1989). The typical velocity of a galaxy is 1000km s � 1

nearthe Virgo Clustercenter. The velocity dispersionof amolecular cloud is typically 1 km s � 1. SincethepressureP is proportional to ρv2, the ram andthe ISM pressuresarecomparableneartheclustercenter. As a result,theISM pres-surewouldincreaseandthehigher ISM pressurewouldmakefmol larger.

We present the Σ— fmol relation in figure 4, adopting�P P0 ��� � Σ Σ0 � 2 � n2. Thesecondtermindicates anincre-

mentof theISM pressuredueto theICM or therampressure.Figure4showsthattheICM pressuremakes fmol muchlarger.Thus,we canconclude thata higherexternal pressureduetoICM or rampressurescouldbetheorigin of thelarge f mol.

3.2.3. Larger Metallicity

Becausethe metallicity strongly affects fmol, the largermetallicity might be an origin of the large fmol. In order toexamine whetherthelargemetallicity makes fmol large,dataon themetallicity distributions arenecessary. However, thereis no availabledatafor thesegalaxies,except for NGC 4254.Therefore,webelow present anew ideatocalculatethemetal-licity distribution usingthe CO, HI, andHα datainsteadofusingthemetallicitydata.

The metallicity is expressedusing the oxygen abundance�12 � log �O H � � , whichis correlatedwith theconversionfac-

tor X (Arimoto etal. 1996). We heredefine

logZ � � 12 � log �O H � � � (5)

andadopttherelationship

log

�X

1020 � ��� logZ � 9 � 3 � (6)

Themetallicityof thesolarneighborhood logZ0 is takento be8.9. We cancalculatetheH2 surfacedensityusingX if Z isgiven. Combining theHI andH2 surfacedensities,we obtainthe total gasdensityΣ, which givesthe pressureP basedonequation(2).

TheUV strengthcanbecalculatedusingtheHα data.Hαimages of the Virgo Clustergalaxies,except for NGC 4402,werearchivedby Koopmann,Kenney, & Young (2001). Fig-ure 5 shows contours of Hα superimposedonto the CO im-ages. We convolved the Hα imagewith the samebeamsastheHI data.WeadoptedU0 � 2 � 42 � 10� 7 erg cm� 2 s� 1 sr� 1

(Reynolds1984). For NGC 4402, we estimatedU while con-sideringtheSchmidtlaw,

U � U0

�ΣΣ0 � n � (7)

wherewe adoptedn � 1 � 33(Komugi et al. 2005).Thus, we can calculateX , P, andU if Z is given. Us-

ing obtainedP, U , and Z, we can calculatethe molecularfraction fmol basedon the modelof Elmegreen(1993). Onthe other hand, we can independently calculate f mol usingonly the HI andH2 surfacedensities.We searchfor an ap-propriate metallicity, which gives the same fmol in the twoways.Thesearching range of themetallicity is adoptedto be6 � 68 � logZ � 11� 12.

Thisprocedureis presentedasaflow-chart in figure6.We plottedthecalculatedmetallicity against theradiusfor

eachgalaxy (figure 7). Eachpoint correspondsto eachob-servedpoint. Themetallicitiesof NGC 4254 andNGC 4654show thatlogZ is about 9.5– 10atthegalacticcenter, andthatit gradually declines.Themetallicity of NGC 4254wascal-culatedby Vila-Costas& Edmunds(1992)andZaritsky et al.(1994). Our obtainedvalues arealmostconsistentwith theirresults.

On the other hand, the metallicities logZ of NGC 4402,NGC4569,andNGC4579oftenexceed10.However, thefor-merresearchshows that thereis no galaxy thatgiveslogZ

�10. In addition, thedispersionsof logZ of NGC 4402, NGC4569,andNGC4579 arelargerthantheothers. Moreover, themetallicitieslogZ of NGC4402 andNGC4569 increasewiththeradius.This tendency is unnaturalbecausethemetallicitylogZ usually decreaseswith the radius (Vila-Costas& Ed-munds1992; Zaritsky et al. 1994). In thecaseof NGC 4579,therearemany pointswhereZ cannot becalculatedwithin thesolutionrange 6 � 68 � logZ � 11� 12in theinnerregion.

4 Nakanishi et al.

TABLE 1OBSERVATIONAL PARAMETER

NGC R.A.(J2000) Dec.(J2000) Morph. BT i P.A. Vhel

h m s d m s mag. deg. deg. km s� 1

4254 12 18 49.61 +14 24 59.6 SA(s)c 10.44 42 68 24054402 12 26 07.45 +13 06 44.7 Sb 12.55 75 90 2344569 12 36 49.82 +13 09 45.8 SAB(rs)ab 10.26 64 23 -2354579 12 37 43.53 +11 49 05.5 SAB(rs)b 10.48 38 05 15204654 12 43 56.67 +13 07 36.1 SAB(rs)cd 11.10 52 125 1039

Col. (1): Galaxy name.Cols. (2) and(3): Central position takenfrom Sofueet al. (2003).Col. (4): Morphological typefrom RC3(deVaucouleurset al. 1991).Col. (5): Total B-bandmagnitude taken from RC3.Cols. (6) and(7): Theinclination angleandthepositionangle (P.A.) taken from Koopmann,Kenney, &

Young(2001)or Phookun,Vogel, & Mundy (1993).Col. (8): Systemicvelocity taken from Kenney & Young(1988)or RC3.

Therefore, thepossibilityof highermetallicity is lessplau-sible, becausean abnormal metallicity distribution is neces-saryto reproducesucha large fmol. However, this possibilitystill cannot beruledout, sincenobody hasdirectly measuredthemetallicitiesof NGC 4402, NGC 4569, andNGC 4579.

3.2.4. Lower UV Radiation Field

BecausetheUV strengthaffects fmol, a lowerUV radiationfield mightbetheorigin of thelarge fmol. In orderto examinethis effect, we plottedthe relationbetweenthe UV strengthandthetotalgasdensity(figure8),excluding thecentral1 kpcto avoid Hα emissionfrom active galactic nuclei. This figureshows thatthereis little differencein theUV strength,U U0,for Σ � 20M � pc� 2, although U U0 is lower in NGC 4569thanin NGC 4254andNGC 4654 in thecaseof Σ

�20M �

pc� 2. Thus, it is difficult to attributethelarge fmol to a lowerUV radiation field, becausea large fmol for Σ � 20M � pc� 2

cannot beexplainedby a differencein theUV strength.

4. SUMMARY

We observedfive Virgo spiralgalaxieswith theNRO 45 mtelescopewith BEARS in the 12CO

�J � 1 � 0� line. Com-

paring the CO datawith the HI datato investigatethe envi-ronmentaleffect, we found that(1) theHI gasdisksof NGC4402, NGC 4569, andNGC 4579, locatednearthecenterofthe cluster, are restrictedwithin the inner disks, (2) the HIclouds cansatisfythecriterionof therampressurestripping,while the molecularclouds cannot, and (3) the Σ– f mol dia-grams of thethreegalaxiesshow anunusually large f mol for

a low total gascolumn density. This large fmol might implythattheHI gasisselectivelystripped.Ontheotherhand,NGC4254 andNGC4654, locatedfar from theclustercenter, haveextendedHI disksandshow thenormal fmol expectedfor fieldgalaxies. Most of theHI gasof thesetwo galaxiesmustnothavebeenstripped, in spiteof thestronglydistortedHI disks.

A higher externalpressuredueto theICM-pressure and/orrampressuremight beanotherpossibilitycausingtheunusu-ally large fmol.

Moreover, we examined whetherdifferencesin the intrin-sic conditions (metallicity andUV) causetheunusuallylargefmol. As a result,we found that anunusuallyhigh metallic-ity is necessaryto explain sucha large fmol, although it stillcannot beruledout.

We also found that the large fmol cannot be explained byonly a differencein theUV radiationfield.

Fromthesediscussions,we notethat the Σ— fmol diagramis a goodtool for investigatingtheenvironmental effect. Anabnormal fmol indicatesthat the ISM suffersclusterenviron-mental effects:rampressurestrippingor higherexternal pres-sure.

We are grateful to the members of the NobeyamaRadioObservatory. We appreciatethatM. Honmakindly providedhis codeto calculate fmol. We would also like to thank T.Namba for his helpwith observationsandreductions. F. E. isfinancially supportedby aresearchfellowshipfrom theJapanSocietyfor thePromotion of Sciencefor Young Scientists.

REFERENCES

Arimoto, N., Sofue,Y., & Tsujimoto,T. 1996,PASJ,48,275Cayatte,V., vanGorkom,J.H., Balkowski, C., & Kotanyi, C. 1990,AJ, 100,

604deVaucouleurs,G., deVaucouleurs,A., Corwin, H. G., Buta, R. J.,Paturel,

G.,& Fouque,P. 1991,Volume1-3,XII, 2069pp.7 figs.. Springer-VerlagBerlin Heidelberg New York,

Elmegreen, B. G. 1993,ApJ,411,170Ferrarese,L., Freedman,W. L., Hill, RobertJ.et al. 1996ApJ,464,568Hidaka, M. & Sofue,Y. 2002,PASJ,54,33Honma,M., Sofue,Y., & Arimoto, N. 1995,A&A, 304,1Kenney, J.D. & Young,J.S.1988,ApJS,66,261Kenney, J.D. P. & Young,J.S.1989,ApJ,344,171Komugi,S.,Sofue,Y., Nakanishi,H., Onodera,S.,& Egusa,F. 2005,PASJ,

57,733Koopmann,R. A., Kenney, J.D. P., & Young,J.2001,ApJS,135,125Kuno,N., Nakai, N., Handa,T., & Sofue,Y. 1995,PASJ,47,745Nakai,N. & Kuno,N. 1995,PASJ,47,761Nulsen,P. E. J.,& Bohringer, H. 1995,MNRAS, 274,1093

Phookun,B., Vogel, S.N., & Mundy, L. G. 1993,ApJ,418,113Phookun,B. & Mundy, L. G. 1995,ApJ,453,154Reynolds,R. J.1984,ApJ,282,191Sanders,D. B., Solomon,P. M., & Scoville, N. Z. 1984,ApJ,276,182Sofue,Y., Koda,J., Nakanishi, H., Onodera, S., Kohno,K., Tomita, A., &

Okumura,S.K. 2003,PASJ,55,1189Sorai,K., Sunada,K., Okumura,S. K., Iwasa,T., Tanaka,A., Natori, K., &

Onuki,H. 2000,Proc.SPIE,4015,86Sorai,K., Nakai, N., Kuno,N., Nishiyama,K., & Hasegawa, T. 2000,PASJ,

52,785Sunada,K., Yamaguchi, C., Nakai,N., Sorai, K., Okumura,S. K., & Ukita,

N. 2000,Proc.SPIE,4015,237Vila-Costas,M. B., & Edmunds,M. G. 1992,MNRAS, 259,121Vollmer, B., Cayatte,V., Balkowski, C.,& Duschl,W. J.2001,ApJ,561,708Zaritsky, D., Kennicutt, R. C.,& Huchra,J.P. 1994,ApJ,420,87

Environmental Effects onVirgoSpirals 5

TABLE 2HI DATA PARAMETER

NGC Synth.BeamSize ∆V ∆TB Referencearcsec2 km s� 1 K

4254 25� ��� 24� � 10.3 0.45 14402 21� � � 17� � 25.0 1.7 24569 17� � � 13� � 25.0 2.8 24579 19� � � 18� � 25.0 2.7 24654 24� � � 25� � 10.3 0.44 3

Col.(1): Galaxyname.Col.(2): Synthesizedbeamsize. Col.(3): Velocity resolution. Col.(4): R.m.s.noisein brightnesstemperature.Col.(5)References.—(1)Phookun,Vogel, & Mundy (1993);(2)Cayatte et al. (1990); (3)Phookun& Mundy (1995)

6 Nakanishi et al.

FIG. 1.— IntegratedCO intensity maps(color) andtheintegratedHI intensity maps(contour). Thecontour levels are100,200,400,800,1600K km s� 1.

Environmental Effects onVirgoSpirals 7

FIG. 2.— Molecular fraction, fmol, againstthesurfacedensity of thetotal gasΣ calculatedusinga formulaof Elmegreen (1993).

FIG. 3.— Molecular fraction, fmol, against thesurfacedensity of the total gasΣ. Thedashedcurves in all panels denotetheaveraged fmol of NGC 4254andNGC 4654.Thedataof ΣHI � 3 8 � 1020 H cm� 2 or ΣH2 � 7 0 � 1019 H2 cm� 2 (3σ) areplotted.

8 Nakanishi et al.

FIG. 4.— Molecular fraction, fmol, againstthe surfacedensityof the total gasΣ calculated usinga formula of Elmegreen (1993)andadopting ! P " P0 #%$! Σ " Σ0 # 2 & n2 (n $ 0, 1, 2, 3, and4).

Environmental Effects onVirgoSpirals 9

FIG. 5.— IntegratedCO intensity maps(color) andtheHα maps(contour). Thecontour levels are2 7 � 10� 16 and2 7 � 10� 15 ergscm� 2 s� 1 arcsec� 2.

10 Nakanishietal.

FIG. 6.— Flow chartto calculate themetallicity usingHI, CO,andHα data.

Environmental EffectsonVirgo Spirals 11

FIG. 7.— Metallic ity logZ $ ! 12 & log 'O" H ( # against the radius. The metallici ties werecalculated using the dataof ΣHI � 3 8 � 1020 H cm� 2 or ΣH2 �7 0 � 1019 H2 cm� 2 (3σ). Thesolid curvesandvertical segments indicatethemeanvaluesandstandarddeviationsat eachradius.

12 Nakanishietal.

FIG. 8.— Relationship between theUV strength ! U " U0 # andsurface densityof thetotal gasΣ.