The Virgo High-Resolution CO Survey. III. NGC 4254

31 Pages·0.15 MB·English

by  Yoshiaki Sofue

The Virgo High-Resolution CO Survey. III. – NGC 4254 –
Yoshiaki Sofue,1 Jin Koda,1,2 Hiroyuki Nakanishi,1 and Makoto Hidaka1
3 1 Institute of Astronomy, University of Tokyo, Mitaka, Tokyo 181-0015
0 2 Nobeyama Radio Observatory, National Astronomical Observatory, Mitaka, Tokyo 181-8588
0
E-mail [email protected]
2

n
(Received 2002 August 15; accepted 2002 October 15)
a
J

1
Abstract
(Phookun et al. 1993). We adopted the Cepheid
1
calibrated distance of 16.1 Mpc of the Virgo Cluster
v
We present high-angular-resolution (1′′.5–5′′) interfer- for this galaxy (Ferrarese et al. 1996), with which
0
1 ometer observations of the 12CO (J = 1–0) emission the Tully–Fisher distance of 16.8 Mpc is consistent
0 in the central region of the SA(s)c galaxy NGC 4254. (Scho¨niger,Sofue1997). Theparametersofthegalaxy
1 The observations were obtained using the Nobeyama are given in table 1. Optical images show that the
0
Millimeter-wave Array (NMA) during the course of a galaxyis dominatedby a three-armstructure, exhibit-
3
long-term CO line survey of Virgo spirals. We present ingm=1andm=3modes(Iyeetal. 1982;Gonz´alez,
0
/ the spectra, channel maps, integrated intensity dis- Graham1996). Thisone-armedstructureisalsoseenin
h tributions, velocity fields, position–velocity diagrams, theHigasdistribution(Phookunetal. 1993). Suchan
p
- andcomparethedatawithvariousopticalimages. The asymmetric spiral pattern is often observed in tidally
o rotation velocity is already finite at the nucleus, or at interactinggalaxies,butthereisapparentlynomassive
tr least it rises steeply to 80 km s−1 within the cen- companionaroundNGC4254. Moreover,atidalinter-
s tral 1′′, indicating the existence of a massive core of actiongenerallyproducesabisymmetricstructurewith
a
: 108M⊙• within 1′′ (80 pc) radius. The CO intensity near and far side arms, and the resultant spiral struc-
v
mapsshowthattheinnerdiskhaswell-developedmul- ture showsm=2mode features,butnotthem=1or
i
X tiple spiral arms, winding out from a bar-shaped elon- m=3mode. Therefore,theasymmetricspiralpattern
r gated molecular complex. In addition to the bisym- ofNGC 4254wouldbe duetosomeothermechanisms,
a
metric spiral arms, an asymmetric tightly wound arm such as an interaction with the intracluster medium
with high molecular gas density is found to wind out (ICM), as suggested by the H i head–tail structure. It
from the molecular bar. The molecular spiral arms, isalsoaninterestingquestionastohowdeeplytheICM
particularly the tightly wound arm, well traces optical rameffectcanaffectthegalacticdisksandwhetherthe
dark lanes, and are associated with Hα arms having molecular disks are indeed disturbed. In fact, Hidaka
many H ii regions. The inner asymmetric spiral struc- andSofue(2002)haveshownbyanumericalsimulation
tures can be explained by ram-pressure distortion of that a ram effect by ICM wind can produce lopsided
inter-arm low density regions of the inner disk by the molecular arms in the central few kpc region, if the
intra-cluster gas wind, and is indeed reproduced by a inter-arm gas density is low enough even though the
hydrodynamical simulation. mean radial density is high enough to avoid the ram
Key words : clusters: individual (Virgo) — galaxies: effect. In such a circumstance, the ICM ram pressure
individual (NGC 4254, M 99) — galaxies: spiral — perturbs the orbits of the low-density inter-arm gas,
galaxies: kinematics and dynamics — galaxies: ISM which results in significantly displaced shocked arms
— ISM: molecular gas from unperturbed arms.
In this paper, we present high-resolution 12CO
1 Introduction (J = 1–0) line observations of NGC 4254 using the
Nobeyama mm-wave Array (NMA) in AB, C and D
TheVirgomemberspiralNGC4254(M99)isanSA(s)c array configurations. The observational parameters
galaxy (RC3) with an optical radius of 5′.36 (de Vau- and detailed procedures are described in the first pa-
couleurs et al. 1991), and an inclination angle of 42◦ perofthisseries,reportingthehigh-resolutionCOline
1
Virgo survey (VCOS: Sofue et al. 2003a). We dis- thus-obtained dirty map by the CLEAN method us-
cuss the kinematics and ISM properties based on the ing three different weighting functions and tapering,
CO data, and compare the result with various opti- as summarized in table 4.
cal images. Environmental effects, such as the tidal First,weobtainedlow-resolutionmapsusinganatu-
interaction and/or the ram-pressure effects, would be ralweightingfunction, taperedby aGaussianfunction
an interesting subject for some Virgo galaxies located havingdeviationsof60kλandacut-offatgreaterthan
near to the cluster center and moving in the dense in- 80kλin(u,v)space. Thedatawereaveragedin4bins,
tracluster medium. Based on the CO data, we discuss yielding a 128-channel data cube with a velocity res-
the possibility that the distribution and kinematics of olution of 21 km s−1 ; the channel increment was 2,
molecular clouds in the central region can be affected correspondingto10.4 kms−1 . Thesynthesizedbeam
by the ram-pressure effect of the intracluster medium. was5′′.22×3′′.45,andtheobtainedchannelhadatyp-
ical r.m.s. noise of 19.6 mJy/beam.
Second, we obtained the most representative maps
2 Observations fromthepresentobservationsusinganaturalweighting
function, no taper. The data were averaged in 4 bins
12CO (J = 1–0)high-angular-resolution interferome- (21 kms−1 ) ofthe originalchannels,andthe channel
ter observations of the central region of NGC 4254 incrementwas2(10.4 kms−1 ),yieldinga128-channel
were carried out on 2000 March 10–11, 2000 Febru- data cube. The synthesized beam was 2′′.99×2′′.33,
ary 4–5, and 1999 December 8, using the Nobeyama andthetypicalr.m.s. noiseonachannelmapwas15.4
Millimeter Array (NMA) in the AB, C, and D con- mJy/beam.
figuration, respectively. Tables 2 gives the observa- Third, we obtained high-resolution maps, using a
tional parameters. Each observation run took typ- uniform weighting function, no taper. The data were
ically 8 hours, including calibrations. The point- averaged in 2 bins giving a velocity resolution of 10.4
ing center was at RA (J2000)=12h18m50.03s, Dec km s−1 ; the channel increment was 2 (10.4 km s−1
(J2000)=+14◦24′52′′.8, adopted from Condon et al. ). The synthesized beam was 1′′.67×1′′.53, and the
(1990). However, this position was found to be 5′′.9 typical r.m.s. noise was 21.3 mJy/beam.
to the east and 6′′.3 to the south of the newly de-
termined kinematical center of the galaxy using our
3 Results
high-resolution CO-line velocity field. In this work,
therefore, we adopted the new kinematical center as
3.1 Spectra
our origin of the maps, whose coordinates are RA
(J2000) = 12h18m49.61s, Dec (J2000)= +14◦24′59′′.1. Figures 1a and b show CO line spectra averaged in a
Thepresentcenterpositionmaybecomparedwiththe 1′ × 1′ squared region (4.89 kpc square) around the
other observations listed in table 3, which all coincide center and a 20′′×20′′ (1.69 kpc) squared region, re-
within ∼±0′′.5. spectively. Both spectra indicate double horn shapes,
We also observed nearby radio point source 3C 273 typical for a rotating disk. However, the central spec-
as a flux and phase calibrator every 20 min. Because trum in figure 1b has a much narrower width, ∼ 130
the intrinsic flux density of 3C 273 at the observing kms−1,thanthatoftheouterregion,wherethewidth
frequency was non-periodically variable,we performed is 210 km s−1 . Since the double-horn feature is typi-
flux calibration for each observation. The flux den- calforarotatingdiskwithaconstantrotationvelocity,
sity of 3C 273 was 11.25, 12.40, and 11.26 Jy, in the the clearly different velocity widths suggest that the
observations of the AB, C, and D configurations, re- diskrotationvelocityvariesstep-likeatafewkpcfrom
spectively. We used a spectro-correlatorsystem, Ultra the center, which is discussed in a more detail using
Wide BandCorrelator(UWBC: Okumuraet al. 2000) position–velocity diagrams in a later section.
in a narrow-band mode, which had 256 channels; the Figure 1c shows a CO line spectrum of NGC 4254
total bandwidth was 512 MHz. One channel corre- obtained by convolution with a synthesized Gaussian
sponded to 5.24 km s−1 at the observing frequency. beam of FWHM 45′′ in order to compare the present
The raw data were calibrated using UVPROC- COintensitywiththatoftheFCRAO14-msingle-dish
II, a first-stage reduction system developed at the observation(Kenney,Young1988). Thepeakintensity
Nobeyama Radio Observatory (NRO), and were then and the integrated intensity of our NMA observation
Fourier-transformed using the NRAO Astronomical are2.39Jy/Beamand325.58Jy/Beam kms−1,corre-
Image Processing System (AIPS). We reduced the spondingto106mKand14.8K kms−1 ,respectively.
2
Table 1: Properties of observed galaxies
Morphology1 SA(s)c
NED position2 (J2000) α = 12h18m49.56s
δ = +14◦24′59′′.4
Apparent magnitude1 10.44
Systemic velocity1 2407 km sec−1
Inclination angle3 42◦
Position angle3 68◦
Assumed distance4 16.1 Mpc (1′′ =78.06 pc)
References: 1. de Vaucouleur et al.
(1991); 2. NASA Extragalactic Database (http://nedwww.ipac.caltech.edu/); 3. Phookun et al. (1993); 4.
Ferrarese et al. (1999).
Table 2: Observation Parameters.
Observed center frequency 114.353GHz
Array configurations AB, C, and D
Observing field center (Condon et al. 1990):
α (J2000) 12h18m50.03s
δ (J2000) +14◦24′52′′.8
Derived dynamical center:
( assumed to be nucleus position ≃ map center in this paper)
α (J2000) 12h18m49.61s
δ (J2000) +14◦24′59′′.1
Central velocity 2400 km s−1
Frequency channels 256
Total bandwidth 512 MHz
Velocity coverage 1342 km s−1
Velocity resolution 5.24 km s−1
Amplitude and phase calibrator 3C 273
Primary beam 65′′
Cell size 0′′.25
Table 3: Comparison of our center position with other observations.
RA (J2000) Dec (J2000) References
12h18m49.61s +14◦24′59′′.1 CO: This work
12h18m49.63s +14◦24′58′′.8 CO: Sakamoto et al. (1999)
12h18m49.43s +14◦24′59′′.5 B-band photometry: Yasuda et al. (1995)
12h18m49.73s +14◦24′58′′.8 IRAS: Soifer et al. (1987)
3
Table 4: Parameters of Maps†.
Beam Velocity r.m.s. noise T for
b
Resolution Weighting FWHM P.A. Resolution Sampling σ 1 Jybeam−1
(arcsec) (deg) (kms−1) (kms−1) (mJybeam−1) (K)
Low Taper 5.22×3.45 152.0 21.0 10.5 19.6 5.1
Medium Natural 2.99×2.34 148.0 21.0 10.5 15.4 13.2
High Uniform 1.67×1.53 167.5 21.0 10.5 21.3 36.0
† The map centers are set at the derived dynamical center:
(α , δ ) =(12h18m49.609s, +14◦24′53′′.1)
2000 2000
A FCRAO CO observation of NGC 4254 shows that 3.2.1 Central CO bar
the peakintensityis89/ηB mK,andthe integratedin-
The CO emission shows an elongated bar-like concen-
tensity is (10.6±2.0)/η K km s−1 , where η is the
B B
tration around the map center (kinematical center)
beamefficiencyofthe FCRAO14mtelescope,whichis
with the major axis at a P.A. of about 50◦, being dis-
equal to 0.53±0.04. Hence, the integrated intensity
placedfromthegalaxy’smajoraxisatP.A.=68◦. The
of our NMA observation covers 74% of the integrated
center of gravity of this bar is slightly offset from the
intensity of the FCRAO single-dish observation.
kinematicalcentertowardthenorthwest. Althoughthe
— Fig. 1a, b, c — CO distribution shows a bar feature, the optical mor-
phology is non-barred SA(s)c type (RC3). In fact, no
bar feature is recognized in either the R-band image
shown in figure 6a or in K-band image in figure 8, as
is shown later.
3.2 CO Intensity Maps
3.2.2 Spiral arms
Figure 2 shows channel maps at a velocity separation
of 10.4 km s−1 . Each channel map displays an aver- Two major spiral arms are winding out from both
aged brightness in a 21 km s−1 velocity range. The ends of this central molecular bar in an counterclock-
CO emission is visible from the 6th channel at V = wise direction, and extend until the field edge. In fig-
lsr
2300 km s−1 to the 27th channel at 2517 km s−1 . ure 3 we name the western major arm Arm I, and
The distributions in individual channel maps are gen- the eastern major arm Arm II. At RA=12h18m51.5s,
erallyalongiso-velocitylines,typicalforarotatingdisk Dec=14◦24′10′′ (J2000),there is a segment of an arm-
galaxy, while they are patchy, indicating that the dis- like feature with a dense molecular complex, running
tributionisnotuniform,butconcentratedinthearms. parallel to Arm II.
— Fig. 2 —
3.2.3 Tightly wound arm
Figure 3 shows the total CO intensity map at a low Soon after it starts fromthe easternend ofthe central
resolution of 5′′.22×3′′.45. Figure 4a shows the total CO bar, Arm II bifurcates into a more tightly wound
CO intensity map at the representative resolution of densearm,whichwecallArmIII,asindicatedinfigure
2′′.33×2′′.99 using a natural weighting function. No 3. The western end of this tightly-wound arm runs
primarybeam correctionhas been applied. The major closely parallel to Arm I, as is more clearly seen in
andminor axesacrossthe dynamicalcenter are shown figures 4a and 5a. This tight arm has a much higher
by the big crossesin figures3 and 4a. Figure 5ashows brightness, about twice to three times that of the two
the same view but at the highest resolutionof 1′′.67× major spiral arms.
1′′.53 with uniform weighting.
3.2.4 North–South Asymmetry
— Fig. 3 —
— Fig. 4a, b, c— Theglobaldistributionofmoleculargasintheobserved
— Fig. 5a, b, c — region of NGC 4254, as shown in figures 3 to 5, is
4
highlyasymmetricwithrespecttothemajoraxis. The 3.4 Overlays on Optical Images
southern half is much CO brighter compared to the
northern half. Most of the asymmetry comes from the
tight CO arm, Arm III. In figure 6a we overlay our CO intensity contour map
at medium resolution (figure 4a) on a g-band image
from the Princeton Galaxies Catalogue obtained with
3.3 Velocity Field
the Palomar 1.5-m telescope by Frei et al (1996). The
kinematical center was assumed to coincide with the
Figure 4b shows an intensity-weighted velocity field
optical nucleus. Scaling and positioning are accurate
correspondingtofigure4awithnaturalweighting,and
within an error of about 2′′, which applies to all the
figure 4c is an overlayof the same velocity field on the
following overlays. Figure 6b shows the same region,
integrated intensity map at a lower resolution in the
but superposed on a 5957A-bandimage from the HST
gray scale, which is the same as figure 3. Figure 5c
WFPC2 archive. Figure 6c shows the central region
shows the velocity field for the central region at high
with the high-resolution CO contour map on the HST
resolution, corresponding to figure 5b. Figure 5d en-
image as in figure 6b. The CO arms, particularly the
larges the central 4′′ region of figure 5c.
tightly-wound dense CO arm (Arm III), well trace the
Wedeterminedthekinematicalcenterastheposition
optical dust lanes running along the inner sides of the
where the iso-contours run most tightly in the high-
optical spiral arms. An inner dust lane at about 6′′
resolutionvelocityfieldinfigures5candd: RA(J2000)
south-east of the nucleus is also traced by a CO arm,
=12h18m49.61s andDec(J2000)=+14◦24′53′′.1with
starting from the eastern mid-point of the central CO
an accuracy of ±0′′.3. We adopt these coordinates
bar. InsidetheCObar,mostoftheCOclumpsaregen-
as the center position of the galaxy, and assume that
erallyassociatedwith dustlanesand/ordustyclumps.
the position coincides with the nucleus. These coor-
However, the molecular bar, itself, does not well draw
dinates are in accordance with the dynamical center
an overallspiral pattern, whereas the HST optical im-
determined by applying the task GAL in the AIPS re-
age within 10′′ of the nucleus shows symmetric amor-
duction package, which uses a Brandt rotation curve
phous spiral patterns.
to fit the (RA, Dec, Velocity) cube. The fitted result
tothe entiredisk intheobservedareawasRA(J2000) Figure 6d shows an overlay of a medium-resolution
= 12h18m49.56s, Dec (J=2000) =+14◦24′58′′.5. The CO map on a negative Hα image taken from the NED
GAL fitting also gave the position angle of the molec- archive of a survey of H ii regions in spiral galaxies
ulardisktobe P.A.=66◦.5,inclination38◦.8,andthe obtained by Banfi et al (1993) using the 2.1-m KPNO
systemicvelocity2403.6 kms−1 ,whichareconsistent telescope. Thisfiguredemonstratesanexcellentglobal
with the adopted parameters in table 1. correlation of molecular arms and clouds with H ii re-
The generalpattern of the velocity field in figure 4b gions, indicating that star formation occurs in spiral
shows a symmetric spider diagram, indicating a regu- arms with dense molecular clouds. A more detailed
lar circular rotation of the CO disk, with the eastern inspection of figure 6d reveals, however, that some in-
half red-shifted and the western half blue-shifted. As- tense molecular clouds are not directly superposed on
suming that the spiral arms are trailing, the rotation H ii regions, but are slightly displaced from the star-
is clockwise and, accordingly, the northern side is the forming region. Also, some strongest H ii regions are
near side. On smaller scales, the non-circular veloc- not superposed by CO clumps, but are slightly shifted
ity components are superimposed. Individual spiral frommolecularcloudcenters. Thedisplacementgener-
arms show iso-velocity contours running obliquely to ally occurs in the direction perpendicular to the arms
thoseexpectedfromacircularrotation,indicatingnon- in the sense that that the H ii regions are located
circular streaming motion due to spiral density waves. outside of the CO arms (dark lanes). This is rea-
The central CO bar, whose major axis is at about sonably explained by the galactic shock-wave theory.
P.A. = 50◦, is rotating rather circularly with the node However, the displacement also occurs among H ii re-
of the velocity field coinciding with that of the P.A. gions and molecular clumps along the arms. This fact
of the galaxy at 68◦. However, the very central re- shows that intense molecular clouds are places where
gion within 2′′ from the nucleus is superposed by non- starformationhasnotyetstarted. Ontheotherhand,
circular motion, where the iso-velocitycontours run in thestrongestHiiregionshavealreadyexhaustedtheir
an integral-sign shape with their average direction in parent molecular clouds.
thenorth-south,andthevelocitynodeisatP.A.=80◦,
significantly displaced from the galaxy’s major axis. — Fig. 6a, b, c, d —
5
3.5 PV diagram shows the obtained rotation curve with an inclination
angleof42◦beingcorrected. Therotationvelocityrises
The position–velocity (PV) diagrams along the major
steeplyinthecentral∼100pc,andreachesto120 km
axisatP.A.=68◦crossingthenucleusofNGC4254are
s−1 maximum, followed by a small dip at 250 pc. It
shown in figures 7a to c in the order of low, medium,
then rises to a maximum velocity of 190 km s−1 at
and high resolutions, respectively. The slit widths are
r ∼ 900 pc, and declines to 150 km s−1 at 1.5 kpc.
10′′, 3′′ and 2′′, respectively. Figure 7c enlarges the
Thismaximummaybeduetothecentralbulge. Then,
central part.
the rotation velocity gradually increases, representing
Emission in the PV diagram is separated into two
thediskcomponent. Thesmall-scalefluctuationofam-
major concentrations around the nucleus at finite ve-
plitudesby∼±10 kms−1 maynotberealand,hence,
locity offsets of about ±40 km s−1 from the systemic
the small local dips should not be taken seriously.
velocity. The PVdiagramshowsa clearintensity min-
The dynamical mass in the central 100 pc is 3 ×
imum at the nucleus, which is prominent in figure 7c,
108M for a rotationvelocity of120 km s−1 , andthe
⊙•
despite the continuous distribution of CO emission in
mass within 1 kpc is 7×109M for 180 km s−1 . The
⊙•
the integrated intensity maps (figures 3 to 5). This
high central velocity and massive core within the 100
indicates that the rotation velocity rises very steeply
pcregionofthenucleuswasobservedformanygalaxies
within the central 1′′ or less.
(Takamiya, Sofue 2000; Sofue et al. 2001; Koda et al.
AfterattainingtheintensitymaximuminthePVdi-
2002).
agram, the velocity increases more gently toward the
edges of the bar. Near the ends of the bar, the PV — Fig. 9a, b —
velocity increases step-like, and attains maximum ve-
locities of about ±100 km s−1 in the spiral arms. We calculatedthe surface-massdistributions(SMD)
using the observed rotation curve by applying a de-
— Fig. 7a, b, c — convolution method described in Takamiya and So-
fue (2000). Figure 9b shows the derived mass dis-
tributions, where the dashed line was calculated by a
4 Discussion
spherical-symmetry assumption, and the full line by
a flat-disk assumption. Since the rotation curve is
4.1 Rotation Curve and Dynamical
given only within 3 kpc, the results beyond 2.5 kpc
Mass
are not real. Both the dashed and full-line results
agree with each other, except for the larger fluctua-
The inclinationangleofNGC 4254wasobtainedto be
tions and slightly larger values for the spherical as-
i=42◦ fromH i data (Phookunet al. 1993). Figure 8
sumption. The figure indicates a high-density central
shows a contour-form K-band image taken from NED
peak,representingthemassivecoreofscaleradiusof80
data archive of the near-IR surface photometry of spi-
pcwiththecentralvaluebeingashighasSMD=(2–3)
ral galaxies (M¨ollenhof, Heidt 2001). We determined
×104M pc−2. Since the values are resolution-limited,
theinclinationanglebyanellipsefittotheiso-intensity ⊙•
the real core radius would be smaller and the density
contours of the K-band image, and obtained an incli-
would be higher. The massive core is followed by an
nationangleofi=42(±1)◦,consistentwiththeearlier
exponentially decreasing part at r ∼ 0.2–1 kpc with
value. Weadoptedthisinclinationangleinthepresent
a scale radius of 600 pc, likely representing the bulge
work. On the other hand, more face-on values have
component. Beyond 1 kpc, the SMD decreases more
been obtained from CO observations: 29◦ (Sakamoto
slowly, representing an exponential disk component of
et al. 1999) or 28◦ (Kenney, Young 1988). Our CO
the scaleradiusofabout3 kpc,thoughthe accuracyis
maps, representing the two open spiralarms,also sug-
much poorer for this component, because of the small
gesta more face-onvalue. However,we may rely more
radius of deconvolution.
on the inclinations from NIR images, because the CO
maps manifest spiral-shocked gaseous arms, but not
4.2 Molecular Gas Mass
necessarily the backbone stellar disk.
The total CO luminosity mass within the inner 22′′.5
— Fig. 8 —
radius region is estimated to be 2.4 ×108M , using a
⊙•
We used PV diagrams to obtain a rotation curve by conversionfactor of C =2.1×1020 cm−2 K−1 km−1 s
applying an iteration method developed by Takamiya (Arimoto et al. 1996). Taking into account the miss-
and Sofue (2002) and Sofue et al (2003b). Figure 9a ing flux of our NMA observation and that the mass of
6
molecular contents including He and other elements is modelandsimulationprocedureisgiveninHidakaand
M =1.36M ,thetotalmassofmoleculargasiscal- Sofue (2002).
gas H2
culated to be 4.34×108M . The CO luminosity mass
⊙•
within the inner 5′′ radiusregion,where a missingflux — Fig. 10a, b, c —
correction is not needed, is 4.0×107M . The lumi-
⊙•
The simulationmay be comparedwith the observed
nosity mass to dynamical mass ratio in the inner 5′′ is
CO intensity distribution, shown in figure 10c (bot-
0.15, much larger than that in the inner 22′′.5, where
tom),thesameasfigure4a. Thedistributionofmolec-
the ratio is 0.026. If we adopt a more face-on value
ular gas in the central 1′ region of NGC 4254 is ex-
forthe inclination,e.g. 29◦,theaboveratiosshouldbe
tended to the south-eastern half, where the tightly
decreasedbyafactorof0.52. Hence,themolecular-gas
wound CO arm is most prominent. This tight arm
mass in NGC 4254 is not particularly large compared
winds out to the north, followed by the NE opti-
with the dynamical mass.
cal outer arms. The western arm winds out to the
south, andcontinues to the optical/Hione-armedspi-
4.3 Origin of the North–South Asym- ral arm. These asymmetric arm structures, particu-
metry in CO larly the tight CO arm (Arm III), are well mimicked
by the simulation in figures 10a and b, although the
Environmental effects, such as the tidal interaction
details are not necessarily reproduced.
and/ortheram-pressureeffects,areaninterestingsub-
ject concerning the Virgo galaxies. NGC 4254, which
We are gratefulto the staffat NRO for their helpful
is apparently not associated with a companion galaxy,
isknownforitsdistortedHistructure,andthelopsid- discussion about the observation and reduction. The
optical images were taken via the NASA Extragalac-
edness is supposed to be the result of an environmen-
tic Data Archive (NED), and from the Hubble-Space-
tal effect due to the ram pressure by the intracluster
medium (Phookun et al. 1993). The outer H i gas is Telescope Data Archive at STScI operated by NASA.
largelyextendedinthenortheasternareaover2′′(∼10 J.K.wasfinanciallysupportedbytheJSPS(JapanSo-
cietyforthePromotionofScience)foryoungscientists.
kpc), and is associated with several bifurcated optical
arms. One prominent optical/H i arm extends from
thesouth-to-westernregion;thisprominentone-armed
feature leads to the m = 1 mode based on a spiral References
mode analysis (Iye et al. 1982). Arimoto, N., Sofue, Y., & Tsujimoto, T. 1996, PASJ,
It was long believed that the inner disks are sta- 48, 275
ble and not disturbed by such an environmental effect Banfi, M., Rampazzo, R., Chincarini, G., & Henry, R.
as the ram pressure (Vollmer et al. 2001). However, B. C. 1993 A&A 280, 373
simulations have already shown that the ram effect is Condon, J. J., Helou, G., Sanders, D. B., & Soifer, B.
significant, even on the inner disks (Sofue 1994). Re- T., 1990,ApJS, 73, 359
cently,weexaminedtherameffectonthe innermolec- de Vaucouleurs, G., de Vaucouleurs, A., Corwin, H.
ular disk in detail, and showed that it is significant G. Jr., Buta, R. J., Paturel, G., & Fouqu´e P. 1991,
when it acts on the inter-arm regions, where the gas Third Reference Catalog of Bright Galaxies (New
density is much lower than in the arm regions, and York: Springer-Verlag)
hencetheram-pressureeasilydisturbstheorbitsofthe Ferrarese, L., Freedman, W. L., Hill, R. J., Saha, A.,
inter-armgas (Hidaka, Sofue 2002). The disturbed or- Madore, B. F., Kennicut, R. C. Jr., Stetson, P. B.,
bitsofinter-armgasleadstoasignificantdisplacement ford, H. C., et al. 1996 ApJ, 464, 568
ofthe shockedarmsfromthe regularbisymmetric arm Frei, Z., Guhathakurta, P., Gunn, J. E., & Tyson, J.
positions, resulting in distorted inner molecular spiral A., 1996, AJ, 111, 174
structures. Figures 10a (top right) and 10b (top left) Gonz´alez,R.A., & Graham,J.R.1996,ApJ,460,651
show the result of our two-dimensional hydrodynami- Hidaka, M., & Sofue Y. 2002,PASJ, 54, 33
calsimulationoframpressureonagasdiskwithspiral Iye, M., Okamura, S., Hamabe, M., & Watanabe, M.
arms. Here, the intracluster wind blows from the west 1982, ApJ, 256, 103
toward the east at an inflowing angle of 45◦ to the Kenney, J. D., & Young, J. S., 1988, ApJS, 66, 261
galactic plane with a wind velocity of 1000 and 1500 Koda, K., Sofue, Y., Kohno, K., Nakanishi, H., On-
km s−1 , respectively, andthe ICM density is assumed odera,S.,Okumura,S.K.,&Irwin,JudithA.2002,
to be 5×10−4 H cm−3. A detailed description of the ApJ, 573, 105
7
Mo¨llenhoff, C., & Heidt, J. 2001, A&A, 368, 16 Figure Captions
Okumura, S. K., Momose, M., Kawaguchi, N., Kan-
zawa, T., Tsutsumi, T., Tanaka, A., Ichikawa, T.,
Suzuki, T., et al. 2000,PASJ, 52, 339 PS figures are available from http://www.ioa.s.u-
Phookun, B., Vogel, S. T., & Mundy, L. 1993, ApJ, tokyo.ac.jp/radio/virgo/
418, 113
Sakamoto,K.,Okumura,S.K.,Ishizuki,S.,&Scoville,
N. Z., 1999, ApJS 124, 403 Fig. 1 (a) Spectrum averaged in the central 60′′
Scho¨niger, F., & Sofue, Y., 1997, A&A, 323, 14
squared area, showing a double-horn shape indicative
Sofue, Y. 1994,ApJ, 423, 207
of a rotating disk.
Sofue, Y., Koda, J., Kohno, K., Okumura, S. K.,
Honma,M.,Kawamura,A.,&Irwin,JudithA.2001 (b)Spectrumaveragedinthecentral20′′ squaredarea,
ApJ, 547, L115 showing a double-horn shape, but with a much nar-
Sofue, Y., Koda, J., Nakanishi, H., & Onodera, S. rower width than in figure 1a, indicative of a rotating
2003bPASJ, in this volume. disk at a slower velocity.
Sofue,Y.,Koda,J.,Nakanishi,H.,Onodera,S.Kohno,
K., Tomita, A., & Okumura, S. K. 2003a PASJ, in (c)SpectrumatthecenterafterconvolvedwithaGaus-
this volume. sian beam of FWHM 45′′.
Soifer, B. T., Sanders, D. B., Madore, B. F., Neuge-
bauer, G., Danielson, G. E., Elias, J. H., Lonsdale, Fig. 2. Channel maps of the CO-line intensity at a
C. J., & Rice, W. L. 1987, ApJ, 320, 238 10.4 km s−1 velocity increment, each indicating the
Takamiya, T, & Sofue, Y., 2000, ApJ 534, 670 CO-line intensity integrated in a 21 km s−1 velocity
Takamiya, T, & Sofue, Y., 2002, ApJ, 576, L15 range. The10th,15thand20thchannelscorrespondto
Vollmer,B., Cayatte,V., Balkowski,C., & Duschl, W. V =2340.2,2392.6,and2444.2 kms−1,respectively.
lsr
J. 2001, ApJ, 561, 708 The intensity unit is the brightness temperature in K.
Yasuda,N., Okamura,S., & Fukugita,M. 1995,ApJS, Contoursaredrawnatevery20%ofthe peakintensity
96, 359 of 3.93 K.
Fig. 3. Integrated CO-line intensity map (the zero-th
moment map) of NGC 4254. The uv weighting func-
tion is natural, And tapered. The synthesized beam
is 5′′.22×3′′.45 (P.A. =152◦). The contour levels are
every10%ofthepeakintensityof85.2K kms−1 (16.6
Jy beam−1 km sec−1). The major and minor axes are
indicated by the cross.
Fig. 4 (a) Integrated CO intensity map. The uv
weighting function is natural, and non–tapered. The
synthesized beam is 2′′.99×2′′.34 (P.A. =148◦). The
central cross indicates the dynamical center and the
majorandminoraxes. Becausetheattenuationdueto
the primary beam pattern is not corrected, the noise
level is uniform in this image. The contour levels are
every 10% of the peak flux of 110.3 K km s−1 (8.33
Jy beam−1 km s−1 ). The r.m.s. noise in the map is
1σ =388 mJy beam−1 km s−1 .
(b) Velocity field of NGC 4254 (first moment map)
corresponding to figure 4a. The synthesized beam is
equal to that of figure 4a. The interval of the iso-
velocity contours is 10 km s−1 .
8
(c) Overlay of the velocity field on the low-resolution 12,15,20,... 30,40,... 100%ofthe peakvalue of7.34
intensity distribution from figure 3 in gray. DN/pixel.
Fig. 5 (a) Integrated CO intensity map of NGC 4254, Fig. 9 (a) Rotation curve for the inner 3 kpc of NGC
with uniform weighting. The synthesized beam is 4254 obtained by the iteration method using the PV
1′′.68×1′′.53 (P.A. = 167◦.5). The contour levels are diagram, as described in Takamiya and Sofue (2002)
every 10% of the peak flux of 205.3 K km s−1 (5.694 and Sofue et al. (2003b).
Jy beam−1 km s−1 ).
(b) Surface mass distribution (SMD) obtained by de-
(b) Enlargement of the central part of figure 5a. The convolutionoftherotationcurveusingthe methodde-
beam size and contour intervals are the same as figure scribed in Takamiya and Sofue (2000). The full line
5a. represents the result for a flat-disk assumption, and
the dotted line for a spherical-symmetry assumption
(c)VelocityfieldofNGC4254(firstmomentmap)cor- of the mass distribution. Note the three components:
respondingtofigure5a. Thesynthesizedbeamisequal thecentralmassivecoreat0–100pcwithascaleradius
to that of figure 5a. The interval of the iso-velocity of80pc, abulgeat0.1–1kpc,andthediskcomponent
contours is 10 km s−1 . at 1 to 2.5 kpc.
Fig. 6(a)OverlayoftheintegratedCOintensitymap, Fig. 10. Two-dimensional hydrodynamical simulation
same as figure 5a, on an R-band negative image of oftheeffectoframpressurebyanintraclusterwindon
NGC 4254 reproduced from Frei et al.’s (1996) galaxy the inner part of a spiral galaxy (Hidaka, Sofue 2002).
catalog. (b) Overlay of the integrated CO intensity
map, same as figure 5a, on the B-band positive image (a) (Top left:) A wind with a gas density of 5×10−4
of NGC 4254, as reproduced from the galaxy-image H cm−3 blows from the west (right) to east (left) at
archive of the Hubble-Space-Telescope. (c) The same an inflowing angle of 45◦ with a wind velocity of 1000
asfigure 6a,but enlargedforthe centralpart. (d) The km s−1 . The rotation of the galaxy is anticlockwise,
same as figure 6a, but overlaid on an Hα image from just like NGC 4254. Note that the faint outer features
Banfi et al. (1993) taken from the NED archive. are artifact due to an edge-reflection effect in the cal-
culation. (b) (Top right:) The same, but with a wind
Fig. 7(a)Position-velocitydiagramoftheCOemission velocityof1500 kms−1 . (c)(Bottom:) ObservedCO
in NGC 4254 from the tapered cube (5′′.22 × 3′′.45 intensitydistributioninNGC4254(sameasfigure4a).
beam, velocity resolution 21 km s−1 ) along a 10′′-
width slit on the major axis (P.A. = 68◦) across the
kinematical center. the contour levels are at 5, 10,
20, .... 100% of the peak value of 1.73 K (0.349 Jy
beam−1). The velocity centroid is 2402.6 km s−1 .
(b) Same as figure 7a, but for natural weighting and
a non–tapered cube (2′′.99×2′′.34 beam, and velocity
resolution 21 km s−1 ) along a 3′′-width slit on the
major axis. The contour levels are at 5, 10, 20, 30, ...
100% of the peak value of 3.23 K (0.248 Jy beam−1).
(c)Sameasfigure7a,butforthe centralpartfromthe
high-resolutioncube (1′′.68×1′′.53beam, andvelocity
resolutionof10.4 kms−1 )ina2′′-widthslitalongthe
major axis. The contour levels are every 10% of the
peak value of 5.46 K (0.152 Jy beam−1).
Fig. 8. K-band image of NGC 4254 in a contour
form, taken from the NED archive of the NIR survey
ofnearbygalaxiesbyMo¨lenhoffandHeidt(2001). The
displayedareais4′×4′aroundthenucleus,and1pixel
corresponds to 1′′. The contours are drawn at 2, 4, ...
9
This figure "fig1a.gif" is available in "gif"(cid:10) format from:
http://arXiv.org/ps/astro-ph/0301010v1