Molecular Gas and Star Formation in Bar Regions of Strongly Barred Galaxy

2018/6/7 @ galaxy evolution workshop 2018

Molecular gas and star formation in bar regions of strongly barred galaxy

Fumiya Maeda ( Kyoto Univ. ) K. Ohta (Kyoto Univ) Suppression of SF in bars 2 • HII regions are seen in spiral arms. • In the bar regions, although the dark lane implies the presence of dust and molecular clouds, there is no (massive) star formation.

NGC1300NGC1300 HST Hα B, V, I - band + Hα filter

https://www.spacetelescope.org/images/opo0501a/ What prevents star formation in bar regions? 3 • The molecular clouds may be destroyed by shock due to the high velocity of the gases relative to the bar structure. (e.g., Tubbs 1982) • Strong shear motion in the bar region may prevent molecular cloud formation (e.g., Athanassoula 1992). • Combination of these (Reynaud & Downes 1998).

Recent studies • Molecular clouds in bar regions are gravitationally unbound (e.g., Sorai et al. 2012.) • The high-speed collision of the clouds in the bar region shortens a gas accretion phase of the cloud cores formed, leading to suppression of core growth and massive star formation (Takahira et al. 2014, 2018) • Most of the molecular gases may be diffuse in bar region. → We must resolve molecular gas into clouds. = High resolution CO imaging of barred galaxies is required. Previous works : intermediate-type bar with SF 4 1.3. GIANT MOLECULAR CLOUDS: STELLAR NURSERIES 15 No. 1, 2010 SFE IN THE BARRED SPIRAL GALAXYNo. ] NGC 4303Wide-ﬁeld 12CO (1–0)• imagingSFE 393 in of bar M83 regions is 2~3 times lower than13 that in arm region. • Too large angular resolution to detect clumps. (a) (b) • Many of targets have an intermediate-type bar and do have star-forming regions associated with the bar or bar-like region.

12 CHAPTER 1. INTRODUCTION 1.3. GIANT MOLECULAR CLOUDS: STELLAR NURSERIES 15 No.M83 1, 2010NGC4303 SFE IN THE BARRED SPIRAL GALAXYNo. ] NGC 4303Wide-ﬁeld 12CO (1–0) imaging 393 of M83 13

(a) (b) NGC4303

Arm SFR Figure 1.3. Image of the barred spiral galaxy M83. This image is based on data acquired with the 1.5-metre Danish telescope at ESO’s La Silla Observatory in Chile, through three filters (B, V, R). Credit: ESO/IDA/Danish 1.5 m/R. Gendler, S. Guisard and C. Thone. Bar log Σ Fig. 14. (a, b) Pseudo color representation of the originaland Ho, 1994). The bar structureM83 is likely responsible for feeding gas to the nuclear regionArm Hα image (a) and the background subtracted Hα image(Lundgren (b). et al., 2004a; Fathi et al., 2008). Offset ridges reside along the stellar bar, and those ridges are associated with shocked gas (Ondrechen, 1985). Both images share same color scaling. (c, d) Same as (a)There and are many mapping observations in CO lines (WiklindBar et al., 1990; Handa et al., 1990; Fig. 14. (a, b) Pseudo color representation of the original (b), but for 24µm images. (e) Image of SFR derivedLord from and Kenney, 1991; Kenney and Lord, 1991; Rand et al., 1999; Crosthwaite et al., 2002; Hα image (a) and the background subtracted Hα image (b). the linear combinations of background subtracted HSakamotoα and et al., 2004; Lundgren et al., 2004b; Muraoka et al., 2009a,b; Hirota et al., 2014). Both images share same color scaling. (c, d) Same as (a) and 9 (b), but for 24µm images. (e) Image of SFR derived from Total mass of HI gas is 7.7 10 M (Huchtmeier and Bohnenstengel, 1981). Total mass of H2 24µm images based on the calibration of Calzetti et al. 2007. × ⊙ the linear combinations of background subtracted Hα and gas is 3.2 109 M (Crosthwaite et al., 2002) or 3.9 109 M (Lundgren et al., 2004b). Momose et al. 2010 × ⊙ Hirota et al.× 2014⊙ 24µm images based on the calibration of Calzetti et al. 2007. Formation of spiral arm and bar structures emission removal here. In the subsequent analysis of SFR emission removal here. In the subsequent analysisThe of formation SFR process of spiral arm structures has been well debated, but consensus has not and SFE ( 5.2), to makelog anΣ independentH2 check, SFR will been reached yet. OneFigure of the 10. most popularvs. theoriesplot of spiral (KS structure plot). Left: is the densityvs. wave theoryat the resolution of 6 ( 500 pc; color) at the scale of Kennicutt (1998)’s Figure 9. Right: the same as the and SFE ( 5.2), to make an independent check, SFR will ΣH2 ΣSFR ΣH2 ΣSFR ′′ be derived§ using the Hα alone in the later section. Figure 10. ΣH vs. ΣSFR plot (KS plot). Left: ΣH vs. ΣSFR at the resolution of 6 ( 500 pc; color) at the§ scale of Kennicutt (1998)’s Figure 9of. Lin Right: and Shu the (1964)left same (see panel, also as but Bertinthe at anda larger Lin 1996). scale. They formulated the spiral features not ∼ Fig. 15. (a) Radial distribution of H2 mass. (b) Same as 2 2 ′′ be derived using the Hα alone in the later section.as a specific collection of stars, but rather a density wave that propagates azimuthally through Using the reference boundaries, background emission in (a), but for SFR derived from the linear combination of ∼ Fig. 15. (a) Radial distribution of H2 mass. (b) SameFigure as 1.6. Left: KS plot for the barred spiral galaxy NGC 4303 at the native resolution of 250 pc and the left panel, but at a larger scale. Using the reference boundaries, background emissionthe galactic in disc. Spiral structures can be self-induced and maintained in a globally stable the 24µm image was estimated. This was made by replac- Hα and 24µm images. Solid and dashed lines∼ indicates Figure 1.6. Left: KS plot for the barred spiral galaxydisk. This NGC theory doeslarge 4303(a), not and meet but ata small winding for the SFRdashes dilemma, native derived thatare is, the if from the fitsresolution stellar for the spiral H lineari isand a material combination CO of (N 2501.56smoothing of pc and resolutionof NRO45 theing each ofand island CARMA500 of pc pixels (taken provided identified from as an discrete Momose unprecedented sources et with al. high 2010,the image SFR Figure with and 10). without Right: background Radial subtraction, distributions respec- of H2 ∼ tively. (c) Same as (b), but for SFE ( SFR / MH2 ). the 24µm image was estimated. This was made byarm, replac- the lifetime of theH armα isand shorter 24 thanµm that images. of its hosting Solid galaxy and (> 1 Gyr) dashed because lines the indicates local background emission thereby, which is determined ≡ 0.04) and for CO-only (N 1.37 0.03), respectively.∼= mass, The± SFR derivedfidelity as from well Hα as+24 a highµm, angular and SFE resolution for the (3 barre.2 spiral250 pc), galaxy M83 (taken from Hirota et al. 2014, smoothing resolutioning each of island500 of pc pixels (taken identified from as discrete Momose sourcesgalactic et with rotation al. curve 2010,the is flat SFR and Figure then with the spiral and 10). arm without would Right:= wind background up into± nothing Radial subtraction, after a few distributions respec- of Hby2 two dimensional surface fitting to the boundary pixels.′′ ∼ large and small dashes are the fits for H i and CO (N 1.56 of NRO45 and CARMA provided an unprecedentedgalactic rotations. highdotted This is inconsistent image line is with from the fact Bigiel that most et nearby al. ( galaxies2008)andisfor have spiral arms, N 0.96Figure 15).which SolidAfter and are critical dashedcreating for the lines thebackground indicates accurate image, measurements the it SFRwas smoothed with of and gas without surface background subtraction. ∼ tively. (c) Same as (b), but for SFE ( SFR / MH2 ). 4.4. Radial distribution of rate and efficiency of star for- indicating that the lifetime of the arm should be longer than 1 Gyr. Moreover, the long lived = ± ′′ = mass,± SFR derivedlocal from background Hα +24 emissionµm, thereby, and SFE which for is determined the barre0.07. spiral Our result, galaxyN M830.67, is (taken closer to from Bigiel≡ Hirota et al. (2008 et), al.density 2014,with and a mass median at filter high with resolution. a kernel size We of 18 discussedto remove SFR andmation SFE 0.04) and for CO-only (N 1.37 0.03), respectively. The fidelity as well as a high angular resolutionspirals (3 can′′. explain2 the250 observed o pc),ffset between the stellar arm and dust lane; a supersonic gas discontinuous edges and any other small scale features. by two dimensional surface fitting to the boundaryflow pixels. into a spiral density wave experiences a shock= as a results of the rapid deceleration, and = ± ∼than to Kennicutt (1998), though it is smaller than the linear quantitatively.Finally, a background Our results subtracted are summarized 24µm image wasas follows. made Figure 15(a), (b) and (c) show radial distributions of dotted line is from Bigiel et al. (2008)andisforN 0.96Figure 15).which SolidAfter and are critical dashedcreating for the lines thebackground indicates accurate image, measurements the it SFRwas smoothedthis with occursof before and gas reaching without surface the minima of the background spiral potential (Fujimoto, subtraction. 1968; Roberts, 1969). H mass, SFR and SFE ( SFR / M(H )) , respectively. Recent numericalcorrelation4.4. simulations Radial withN a distributionlive1. (i.e. time-dependent) of rate stellar and disk, however, efficiency shows of star for- by subtracting the smoothed background image from the 2 2 ′′ Right panel1. CO of emission Figure is detected 1.5 shows over the the entire KS plotdisk, i.e., for12 almost nearbyCO (1–0) ev- integrated disk galaxies, intensity≡ is converted not including into H mass high z = ± with a median filter with a kernel size of 18 tothat remove stars and gas populate the spiral arms= for the lifetime of the arm itself, rather than flowing original image. Figure 14a and 14b show the original and 2 0.07. Our result, N 0.67, is closer to Bigiel et al. (2008), density and mass at high resolution. We discussed SFROur andmation data SFE lie roughly in the range of Kennicutt (1998;gray 20 2 into and out of the arm as in the density wave theory (Wada et al., 2011; Baba et al., 2013; erywherebackground including removed 24µ inter-armm images, respectively regions andwith same the downstreamusing the conversion factor of XCO =2 10 cm− (K discontinuous edges and any other small scale features. region), although some data points deviate from the exactstarburst range. galaxies, done by Leroy et al. (2013). The data1 1 is from the HERA× CO-Line Extra- than to Kennicutt (1998= ), though it is smaller than the linear quantitatively. Our results are summarized asGrand follows. et al., 2013). The stellar arms are non-steady (i.e. transient and recurrent); they are sidecolor of scales. the bar. There are remarkable concentrationskm s− along)− (Strong & Mattox 1996, Dame et al. 2001). Finally, a background subtracted 24µm image waswound made and stretchedThe byFigure the deviations galactic 15(a), shear and occur (b) merge and onwith other the (c) arms. spiral show This arms, isradial due to a where distributionsswing the offsetsgalactic of Survey (HERACLES, first maps in Leroy etSFR al. is derived 2009) with used the above the equations, IRAM both 30 for m the telescope the4.3. offset Calculation ridges of of star the formation bar and rate in the ring structurebackground in the subtracted and unsubtracted images. Both correlation N 1. by subtracting the smoothed background image fromamplification the mechanismbetweenH2 mass, that reinforces the SFR gas density and and enhancementstar-forming SFE ( whichSFR regions seeds / aM wake are(H in2 thelarge)) disk , respectively. (Section 4.4). Right panel1. CO of emission Figure is detected 1.5 shows over the the entire KS plotdisk,(Goldreich i.e., and for Lynden-Bell,12 almost nearby 1965b; ev- Julian disk and Toomre, galaxies, 1966;≡ Toomre, 1981; Baba not et al., 2013). including(supported high bynucleus France,z SFR is (r derived Germany,1.6 using kpc the area). and the background The Spain). gas in subtracted the The spiral commonradial arms distributions ex- spatial of H resolution2 and SFR shows of two the peaks survey at the is 1 = original image. Figure 14a and 14b show the originalThis non-steady and stellarAtCO spiral the (1–0) makes 500 the pc integrated gas resolution, spirals associated intensity the with theKS stellar is law converted arms is non-steady. present, into but H2 suffersmass Hα and 24µ∼m images with an SFR calibration presented center and at the end of the bar. Radial distribution of ∼ Our data lie roughly in the range of Kennicutt (1998;gray The timescale of the change of the stellar spiral arm is 1-2 rotational periods at each radius 20 kpc2 sufficienttends to place from many the end resolution of the offset elements ridges toward across the a outer typical disk galaxy. They found a first- erywherebackground including removed 24µ inter-armm images, respectively regions andwith same the downstreamslightlyusing the from conversion the offset factor effect. of TwoXCO types=2 of scatter10 cm appear− (K in by Calzetti et al. (2007). With the SFR calibration of SFE also shows a dip when diffuse background is removed. starburst galaxies, done by Leroy et al. (2013). The data1 1 is from the HERA× CO-Line Extra-region.Calzetti The et al. surface (2007), extinction densities corrected in the Hα outerluminosity spiral arms and region), although some data points deviate from the exact range. color scales. Figureskm s− )10−(a)(Strong and (b). & One Mattox is the 1996, scatter Dame of NGC et 4303 al. 2001).order itself, linear correspondence between ΣSFR and Σmol5.(KS Verifications law) but of the also observational important signs second-order of the side of the bar. There are remarkable concentrations along offset(L(Hα ridgescorr)) is expressedare similar as a linear at high combination resolution. of the Hα The deviations occur on the spiral arms, where the offsetsgalactic Survey (HERACLES, first maps in Leroy eti.e.,SFR al. its is deviation derived 2009) from with used the the average above the of equations, IRAM galaxies, and both 30 the for other msystematic the telescope is 2. variations Spatialand 24µ offsetsm luminosities: about between this H scaling.α and CO This peaks result exist along suggestsstationary the densityan environmental waves dependence of the4.3. offset Calculation ridges of of star the formation bar and rate in the ring structurewithinbackground in the the disk subtracted of NGC 4303. and The unsubtracted distribution images. of our points Both is star formationspiral activitiesL(H arms.αcorr)= HLα between(Hemissionα) + (0.031 internalis0 seen.006)L at(24 the regionsµm) downstream, (4) of aAs nearbyside was summarized of disk in galaxy.1, to what extent the conven- between the gas and star-forming regions are large (Section 4.4). ± § (supported bynucleus France,SFR is (r derived Germany,1.6 using kpc the area). and the background The Spain). gas in subtracted the The spiral commonconcentratedradial arms distributions ex- spatial in a small of region, H resolution2 and and SFR the shows scatter of two within the peaks the survey galaxyat the is gas1where flow,L(H whileα) and theL(24 COµm are emission Hα and 24 isµ upstreamm luminosities, of thetional gas density flow. wave theory and galactic shock model ex- At the 500 pc resolution, the KS law is present, but suffers ∼ iscenter smaller and than at that the of end Kennicutt of the bar. (1998 Radial). Therefore, distribution the scatter ofLeft panelrespectively. of Figure Then, 1.6 it translates shows into the SFR with KS by plot forplain the the real barred aspects of spiral galaxies is galaxy still under NGCactive de- 4303 at Hα and 24µm images with an SFR calibration presented ∼The delay of star formation from the formation ofbate: GMC studies on to test two observable influences of the den- tends from the end of the offset ridges towardamong the galaxies outer seems dominant. Overall, NGC 4303 fits on the 1 42 1 kpc sufficient toby place Calzetti many et al. (2007). resolution With the SFR elements calibration across of SFE a also typical shows a dip disk when digalaxy.ffuse background They is removed. foundthe smoothing a first-spiral resolutionSFR armsM yr would− of=5 cause.3 50010− suchL pc(H offsets.α observedcorr) ergs s− . by(5) Momosesity wave/galactic et al. shock, (2010). namely, the They ’age gradient’ quantified and the slightly from the offset effect. Two types of scatter appear in plot of Kennicutt (1998), but has slightly higher SFE than the ⊙ ∼× region.Calzetti The et al. surface (2007), extinction densities corrected in the Hα outerluminosity spiral arms and differences in3. The starFigure azimuthal formation 14(e)! shows" averaged the activity map of SFR SFE derived between decreases! with the the steeplyabove" galacticenhancement from the regions the of star (nucleus, formation have bar, often spiral revealed arm, and Figures 10(a) and (b). One is the scatter of NGC 4303order itself, linear correspondence between ΣSFR and Σmolaverage.(KS law) but also important second-orderequations, using the background removed Hα and 24µm compelling results (e.g., Cepa & Beckman 1990; Egusa et offset(L(Hα ridgescorr)) is expressedare similar as a linear at high combination resolution. of the Hα 5. Verifications of the observational signs of the circumnuclear disk to the bar and increases towardal. 2004; Tamburro the et al. 2008; Egusa et al. 2009; Foyle et Both Σ and Σ show one order of magnitude scattersinter-arm) in theimages. context of the KS law and found that the SFRs and SFEs are twice as high i.e., its deviation from the average of galaxies, and the othersystematic is variationsand 24µm luminosities: about this scaling. This result suggestsstationaryH2 density anSFR environmental waves dependencespiral of arms. The comparison of SFE in the baral. and 2010; spiral Foyle et al. 2011; Silva-Villa & Larsen 2012; Louie 2. Spatial offsets between Hα and CO peaks existin along Figure 10 the(b), which is consistent with the results ofin Bigiel the spiral armsarms shows than that in SFE the is bar about region. twice as Right high in panel the arms of as Figure 1.6 also shows the difference within the disk of NGC 4303. The distribution of our points is L(Hα )=L(Hα) + (0.031 0.006)L(24µm), (4) et al. (2008, their Figure 4). Therefore, an order of magnitude star formationspiral activities arms.corr Hα betweenemission internal is± seen at the regions downstream of aAs nearby side was summarized of disk in galaxy.1, to what extent the conven-of star formationthose of activity in the bar. (SFRs and SFEs) between the galactic regions in a barred galaxy concentrated in a small region, and the scatter within the galaxy scatter of star formation activity§ is common at a sub-kpc scale. 4. Extreme Σ and SFE are found in the spiral arms, but not gaswhere flow,L(H whileα) and theL(24 COµm are emission Hα and 24 isµ upstreamm luminosities, of theLeroytional gas et density flow. al. (2008 wave)showedthatSFEvariesstronglywithlocal theory and galactic shock modelobserved ex- by Hirota etSFR al. (2014). They observed the nearby barred spiral galaxy M83, which is is smaller than that of Kennicutt (1998). Therefore, the scatterLeft panelrespectively. of Figure Then, 1.6 it translates shows into the SFR with KS by plot forplain the the real barred aspects of spiral galaxies is galaxy still under NGCactive de- 4303 atin the bar, indicating that the trigger of star formation is The delay of star formation from the formation ofconditions. GMC on Since the KS law breaks down at the scalethe of modelling a related object not for only our to the galaxy amount simulations, of available gas, and but found also to the radial dependence of the SFR 1 42 1 fewbate: 100 studies pc, we to could test two attribute observable the variation influences of SFE of the to den-local among galaxies seems dominant. Overall, NGC 4303 fitsthe on the smoothing resolutionSFR M yr− of=5.3 50010− L pc(Hα observedcorr) ergs s− . by(5) Momose et al. (2010). They quantifiedand SFE in thethe bar environment, region: a such low as star galactic formation dynamics activity around spiral in the end of the bar region and, on the spiral arms⊙ would cause× such offsets. environmentssity wave/galactic at the shock, scale of namely, a few 100 the ’agepc. One gradient’ of the and key plot of Kennicutt (1998), but has slightly higher SFE than the ∼ other hand, aarms high and activity the bar. in The the presence middle of of the the active region. star-forming Muraoka et al. (2007) also observed the 3. TheFigure azimuthal 14(e)! shows" averaged the map of SFR SFE derived decreases! with the steeplyabove" environmentalenhancement from the factors the of might star formation be galactic have dynamics often (e.g., revealed spiral differences in star formation activity between the galacticcompelling regions results (e.g., (nucleus, Cepa & Beckman bar, 1990; spiral Egusa et arm, andregions along the spiral arms confirms the visual impression average. equations, using the background removed Hα and 24µm arms). M83 and foundthat the star higher formation SFEs is more in active the nuclear in spiral arms, starburst or reduced region than those in the disk regions. circumnuclear disk to the bar and increases towardal. 2004; Tamburro the et al. 2008; Egusa et al. 2009; Foyle et Both and show one order of magnitude scattersinter-arm) in theimages. context of the KS law and found thatWeverifyascatterofNGC4303itselfamongnearbygalaxies, the SFRs and SFEs are twiceWhy as does highsignificantly the star formation in bar. activity differ depending on the galactic structure’s different ΣH2 ΣSFR spiral arms. The comparison of SFE in the bari.e.,al. and 2010; a scatter spiral Foyle from et the al. KS 2011; law. Silva-Villa Rough estimation & Larsen of scatter 2012; Louieshows in Figure 10(b), which is consistent with the results of Bigiel environments?5. The This SFE derived question with a is metallicity-dependent key to understandingXCO does the not galactic-scale star formation and in the spiral armsarms shows than that in SFE the is bar about region. twice as Right high in panel thethat arms our of results as Figure are a factor 1.6 of 5higherthanthatofKennicutt also shows the differencechange the conclusion, i.e., higher SFE in the spiral arms et al. (2008, their Figure 4). Therefore, an order of magnitude (1998). However, the KS plot∼ has a 1orderofmagnitudehas been the focus of my doctoral research. To understand this, it is important to investigate of star formationthose of activity in the bar. (SFRs and SFEs) between the galactic regions± in a barred galaxythan in the bar, since the difference between the bar and scatter of star formation activity is common at a sub-kpc scale. scatter, as we discussed in Section 5.1.Eventhoughourdatahow the formationthe spiral and arms evolution is reduced of by giant only around molecular 30%. However, clouds (GMCs) is affected by the galactic observed by4. Hirota Extreme etΣSFR al.and (2014). SFE are They found observed in the spiral the arms,are nearby the but fit notby Kennicutt barred (1998 spiral), they are galaxy still within M83, the scatterstructures. which ThisSFE is is in because the circumnuclear the GMCs regions are is a the factor star of two formation to three spots in a galaxy. Leroy et al. (2008)showedthatSFEvariesstronglywithlocal in the bar, indicating that the trigger of star formation(gray region). is higher than the results with the SFE derived by a standard X , since metallicity of the circumnuclear is significantly conditions. Since the KS law breaks down at the scalethe of modelling a object for our galaxy simulations, and found the radial6. SUMMARY dependence of the SFRCO related not only to the amount of available gas, but also to high. few 100 pc, we could attribute the variation of SFE toand local SFE in thethe bar environment, region: a such low as star galactic formation dynamics activity aroundWe observed spiral in the the end barred spiralof the galaxy bar NGC region 4303 in the1.3 and,12CO on Giant6. the The KSmolecular law appears to clouds:break down at stellar our highest nurseriesspatial environments at the scale of a few 100 pc. One of the key (J 1 0) line with NRO45 and CARMA. The combination resolution ( 250 pc); due to the spatial offsets, we find other hand, aarms high and activity the bar. in The the presence middle of of the the active region. star-forming= Muraoka− et al. (2007) also observed the ∼ environmental factors might be galactic dynamics (e.g., spiral regions along the spiral arms confirms the visual impression Giant molecular clouds (GMC) are star formation spots in a galaxy, which makes them an arms). M83 and foundthat the star higher formation SFEs is more in active the nuclear in spiral arms, starburst or reduced region than those in the diskimportant regions. clue for understanding the galactic-scale star formation. They are formed from the WeverifyascatterofNGC4303itselfamongnearbygalaxies,Why doessignificantly the star formation in bar. activity differ depending on the galactic structure’scold di phasefferent of the interstellar medium (ISM), and the densest pockets within them are the i.e., a scatter from the KS law. Rough estimation of scatter shows birth place of stars. Properties (e.g. mass, size, and velocity dispersion) of the GMCs control environments?5. The This SFE derived question with a is metallicity-dependent key to understandingXCO does the not galactic-scale star formation and that our results are a factor of 5higherthanthatofKennicutt change the conclusion, i.e., higher SFE in the spiral arms (1998). However, the KS plot∼ has a 1orderofmagnitudehas been the focus of my doctoral research. To understand this, it is important to investigate ± than in the bar, since the difference between the bar and scatter, as we discussed in Section 5.1.Eventhoughourdatahow the formationthe spiral and arms evolution is reduced of by giant only around molecular 30%. However, clouds (GMCs) is affected by the galactic are the fit by Kennicutt (1998), they are still within the scatterstructures. ThisSFE is in because the circumnuclear the GMCs regions are is a the factor star of two formation to three spots in a galaxy. (gray region). higher than the results with the SFE derived by a standard X , since metallicity of the circumnuclear is significantly 6. SUMMARY CO high. We observed the barred spiral galaxy NGC 4303 in the1.312CO Giant6. The KSmolecular law appears to clouds: break down at stellar our highest nurseriesspatial (J 1 0) line with NRO45 and CARMA. The combination resolution ( 250 pc); due to the spatial offsets, we find = − ∼ Giant molecular clouds (GMC) are star formation spots in a galaxy, which makes them an important clue for understanding the galactic-scale star formation. They are formed from the cold phase of the interstellar medium (ISM), and the densest pockets within them are the birth place of stars. Properties (e.g. mass, size, and velocity dispersion) of the GMCs control Strongly barred galaxies : ideal lab 5

CO observations of strongly barred galaxies are important, because the mechanism of the suppression is expected to be clearly seen in such bars.

NGC1300 is the nearest strongly barred NGC1300 (d = 20 Mpc) galaxy which has remarkable dark lanes in the bar region without HII regions.

• GMCs (~ 50pc) are detectable with ALMA.

However... • CO emission line has not been detected towards the bar regions. 7 野辺山45-mでのCO(1−0)観測（2016, 2017） 15 • ビーム（HPBW） = 13.5” 45-m 野辺山 6 BØNGCefore 1300 going to ALMA: Ø 45NGC-m 5383 observations

• ビNROーム 45=- m1.3 (2016/17) kpc • ビーム = 2.1 kpc • 観• beam測領域 size = 13.”5φ = 1.3 kpc • 観測領域 • Bar A, B Hα なし • Bar A, B HαなNROし 45-m • on source time = 0.5 ~ 2.7hrs / point • Arm A, B Hαあり • Arm A, BHαあり • •rmsArm= 8C ~ 16Hα mKなしat 10 km/s bin 観測領域 NGC5383:Hα NGC1300:Hαα ------dustdust lane lane

40” Arm B

-19°24’00” Arm C Arm A

20” Declination (J2000)

40” Bar B Bar A r = 1.5kpc 3h19m40s 38s 36s 34s Right Ascension (J2000) Maeda et al. 2018 Sheth, et al. 2000, APJ, 632, 217より改変

BarA = Bar1 ArmA = Arm3 BarB = Bar3 ArmB = Arm1 Bar End A = Bar End ArmC = Arm4 7 Molecular観測領域 gases exist in bar regions 2 Maeda et al. 2018 Mmol (Msun) = 4.36 L’CO (K km/s pc ) NGC 1300 Bar A Arm B 40” ]

mK Arm B -19°24’00” Arm C Arm A Intensity [ Intensity 20” Bar B Declination (J2000)

40” Bar A velocity [km/s] velocity [km/s]

3h19m40s 38s 36s 34s Bar B Right AscensionArm (J2000) C Arm A ]

mK BarA = Bar1 ArmA = Arm3 BarB = Bar3 ArmB = Arm1 Bar End A = Bar End ArmC = Arm4 Intensity [ Intensity

velocity [km/s] velocity [km/s] velocity [km/s] 1.2.1 Spiral galaxies (1.2.1)

and Σ H 2 vs. SFR .4 . Σ 1 2 8 very low SFEs in stronglyΣ gas bar region Σ H ∝ vs. SFR µm IR ﬂuxes and Σ Σ SFR Using continuum subtracted Hα image (1.2.2) from HST, we derived SFEs on 1.3 kpc Arm A 1.2 Starscale: formation in galaxies emission ⌃SFR 1 α galaxies and foundSFE a global = relation between(yr the) surface density of SFR and that of gas as ⌃H2 Arm B

1). Figure 1.4(a) shows2 the tight relation of • SFEs in Arm A & B are comparable to 0. . Bar B 0 2 − 1. ± previousThis relationobservations is well known of as thespirals. Kennicutt-Schmidt (KS) relation. Σ H 2 Recent high-resolution) at the observations resolution of have 750 investigated pc in a sample the ofKS 18 relation nearby at non-barred sub-kpc resolution spiral galaxies (e.g., by 2 ∝ +H Σ SFR Arm C Σ HI • SFEs in barBigiel of et NGC1300 al. 2008; Schruba are et ~10 al. 2011). times Figure 1.4 shows the KS relations ( vs. mag.Bar They A investigated how the index of this 1 Σ SFR 1. smaller than that in arm. = Bigiel et al. (2008). SFRs were measured from a combinationAV of FUV and 24- ) gases were measured from CO( (b → SF is remarkablyH 2 suppressed in with a best-ﬁt molecular KS relation of strongly bar regions. Red : Arm , Blue : Bar • The low SFE in Arm C is consistent with Color regions by Bigiel et al. (2008) They reported that these results doshow not change SFEs when in the18 SFRs nearby were measured spiral fromgalaxies H the absent of HII regions there. on 0.75 kpc scale. assuming the average dust extinction of power law change with spatial resolution and showed that the best-ﬁt index hardly changes from ) (a gases were 2 conversion 2

µm IR ﬂuxes and H

8 and a constant CO-to-H 0. of nearby spiral= galaxies at the resolution of 750 pc by 0 H 2 ) + 1− Σ HI I CO( vs. 1)/ 2− Σ SFR CO(

and (b) H 2 Σ 1 . The sensitivity limit of SF tracer is indicated by a horizontal dotted line. vs. 1 )− − 7 Σ SFR 1) with a intensity ratio I − 2 (K km s − 20 cm Figure 1.4: (a) 10 Bigiel et al. (2008)..0 × SFRs were measured from a combination of FUV and 24- measured from CO(2 factor of 2 The vertical dotted line in panel (a) represent the sensitivity limit of the CO emission and the vertical dashed line in panel (b) indicate the transition from the low-density regime to the high-density regime (see text). 1.4 Aim of my studiesNGC5383 9

The very low SFEs are also seen in bar regions of strongly barred galaxy NGC5383.

NGC 5383 (d = 32Mpc) CO(1-0) observations with 45-m

Bar 1 Arm 1

NGC5383 5

• RA = 13h57m04.9s, Dec = +41d50m46 • りょうけん座(CVn)

• vLSR = 2270 km/s , 35Mpc • f = 114.360 GHz Arm 2 RF Bar 2 • 1” ~ 175pc Arm 2 Figure 1.10: SDSS gri image of NGC 5383. This is a strongly barred galaxy with very little star formation RA Dec associated with its two predominant straight dust lanes in the bar region. Bar1 13h57m06.4s 41d50m40.9s Bar 1 Bar 2 Bar end 13h57m08.1s 41d50m23.3s Arm1 13h57m07.2s 41d50m08.0s beamsize = 2.1kpc Bar2 13h57m03.1s 41d50m59.6s Arm2 13h57m02.7s 41d51m24.3s Arm 1 Maeda et al. 2018

Continuum-subtracted Hα image (Sheth et al. 2000)

Sheth et al. (2000)

Bar II

Arm I

Bar I

Arm II

Figure 1.11: Line profiles of the gases in the bar regions and the arm regions in the simulation by Fujimoto et al. (2014a) and Fujimoto et al. (2014b). The observed regions are shown in Figure 1.7(b). Each region is a circular region with a radius of 1.0 kpc and the viewing angle is 35◦ which is the same as that of NGC1300. 1 The width of each bin is 15 km s− .Offset of the velocity is just reflection of the velocity field (rotation curve).

15 1.4 Aim of my studiesNGC5383 10

The very low SFEs are also seen in bar regions of strongly barred galaxy NGC5383. ●：NGC 1300 NGC 5383 (d = 32Mpc) ■：NGC 5383 Arm

NGC5383 5

• RA = 13h57m04.9s, Dec = +41d50m46 • りょうけん座(CVn)

• vLSR = 2270 km/s , 35Mpc Arm 2 Bar • fRF = 114.360 GHz • 1” ~ 175pc Figure 1.10: SDSS gri image of NGC 5383. This is a strongly barred galaxy with very little star formation RA Dec associated with its two predominant straight dust lanes in the bar region. Bar1 13h57m06.4s 41d50m40.9s Bar 1 Bar 2 Bar end 13h57m08.1s 41d50m23.3s Arm1 13h57m07.2s 41d50m08.0s beamsize = 2.1kpc Red : Arm , Blue : Bar Bar2 13h57m03.1s 41d50m59.6s Color regions by Bigiel et al. (2008) Arm 1 Arm2 13h57m02.7s 41d51m24.3s show SFEs in 18 nearby spiral galaxies on 0.75 kpc scale. Continuum-subtracted Hα image

Sheth et al. (2000)

Bar II

Arm I

Bar I

Arm II

15 (c)12CO(1-0) CO(1-0) observations with ALMA 11

ALMA Cycle 5 (2017.0.00248.S)

FoV ~ 54”φ = 3.2 kpc × 2 points (Arm & Bar)(a) Hα

• beam size ~ 0.”42 × 0.”30 = 42pc × 30pc

• on source time ~ 2.7hrs / point Hα ALMA-Arm • rms = 0.7 mJy/beam at 2.0 km/s bin

Arm B

Arm C Arm A (b) B,V,I,Hα

Bar B Bar A

ALMA-Bar (c)12CO(1-0)

12 CO imaging with ALMA (a) Hα

Hα ALMA-Arm We detected the GMC-like gas clouds

Arm B

in the Arm and Bar regions. Arm C Arm A (b) B,V,I,Hα

Bar B Bar A

Velocity integrated map (1550 - 1700 km/s) ALMA-Bar Bar region Arm region

100pc beam 100pc

beam

3.2 kpc Hα + CO map 13

• Arm region : HII regions and GMC-like gas clouds are coexist (e.g., M51). • Bar region : There are GMC-like gas clouds without HII regions. → A possibility that GMCs are formed but massive SFs are not occurred.

Bar region Arm region 14 Comparison観測領域 with 45-m observations

NGC 1300 Bar A Arm B 40”

Arm B -19°24’00” Arm C Arm A 20” Bar B Declination (J2000)

40” Bar A 100pc

3h19m40s 38s 36s 34s Bar B Right AscensionArm (J2000) C Arm A

BarA = Bar1 ArmA = Arm3 BarB = Bar3 ArmB = Arm1 Bar End A = Bar End ArmC = Arm4 Comparison with 45-m observations 15

We estimate the amount of diffuse emission distributed on the scale larger than 400 pc ( = LAS of this ALMA observations). L’co(45m) L’co(ALMA) fraction (K km/s pc2) (K km/s pc2) (%)

Arm A 6.9×106 4.1×106 ~40

Arm B 4.1×106 5.7×105 ~85

Arm C 4.6×106 3.4×105 ~90

Bar A 3.9×106 1.1×106 ~70

Bar B 6.7×106 1.8×106 ~70

• In Arm A, ALMA has detected 40% of total-flux with NRO 45-m.

cf. M51: L’CO(PdBI) /L’CO(PdBI+30m) ~ 50 % (Pety et al 2013)

• In Bar, Arm B & C, diffuse gases may be account for the most of total gases.

→This is the cause of the low SFE in bar regions? Next Step : identification of GMCs 16

We are identifying GMCs in NGC1300 by using clumpfind/CPROPS.

Deriving the properties of GMCs

• Radius → Mvir • velocity width virial parameter • LCO → Mmol • relative velocity distributions

→ Investigation of the cause of SF suppression.

• Gravitationally unbound cloud in bar regions

→ Large virial parameters of GMCs in bar regions

• The high-speed collision of the clouds

→ High relative velocity of GMCs in bar regions Summary 17

• Strongly barred galaxy, NGC1300 is one of the suitable targets to investigate the cause of star formation suppression in bar regions. • However, CO emission line has not been detected towards the bar regions.

CO observations with the Nobeyama 45-m telescope • The molecular gases do exist in the bar regions without Arm associating clear HII regions. • SFEs in bar of NGC1300 are ~10 times smaller than Bar that in arm. CO observations with the ALMA • We detected the GMC-like gas clouds in the arm and bar. • Molecular gas may be mostly distributed on scales larger than 400 pc. • We will identify GMCs and investigation of the cause of SF suppression.