Internal structure of molecular gas in a main sequence galaxy with a UV clump at z = 1.45
Kaito Ushio, Kouji Ohta, Fumiya Maeda, Bunyo Hatsukade, Kiyoto Yabe
DDraft version January 22, 2021
Typeset using L A TEX twocolumn style in AASTeX62
Internal structure of molecular gas in a main sequence galaxy with a UV clump at z = 1 . Kaito Ushio, Kouji Ohta, Fumiya Maeda, Bunyo Hatsukade, and Kiyoto Yabe Department of Astronomy, Kyoto University, Kitashirakawa-Oiwake-Cho, Sakyo-ku, Kyoto 606-8502, Japan Institute of Astronomy, Graduate School of Science, University of Tokyo, 2-21-1 Osawa, Mitaka, Tokyo 181-0015, Japan Kavli Institute for the Physics and Mathematics of the Universe (WPI), University of Tokyo, Kashiwa, Chiba 277-8583, Japan (Received January 22, 2021; Revised January 22, 2021; Accepted January 22, 2021)
Submitted to ApJABSTRACTWe present results of sub-arcsec ALMA observations of CO(2-1) and CO(5-4) toward a massive mainsequence galaxy at z = 1 .
45 in the SXDS/UDS field, aiming at examining the internal distribution andproperties of molecular gas in the galaxy. Our target galaxy consists of the bulge and disk, and has aUV clump in the HST images. The CO emission lines are clearly detected and the CO(5-4)/CO(2-1)flux ratio ( R ) is ∼
1, similar to that of the Milky Way. Assuming a metallicity dependent CO-to-H conversion factor and a CO(2-1)/CO(1-0) flux ratio of 2 (the Milky Way value), the molecular gasmass and the gas mass fraction ( f gas = molecular gas mass / (molecular gas mass + stellar mass)) areestimated to be ∼ . × M (cid:12) and ∼ .
55, respectively. We find that R peak coincides with theposition of the UV clump and its value is approximately two times higher than the galactic average.This result implies high gas density and/or high temperature in the UV clump, which qualitativelyagrees with a numerical simulation of a clumpy galaxy. The CO(2-1) distribution is well represented bya rotating disk model and its half-light radius is ∼ . f gas decreasesfrom ∼ . ∼ . × half-light radius, indicating that the molecular gasis distributed in more central region of the galaxy than stars and seems to associate with the bulgerather than the stellar disk. Keywords: galaxies: evolution — galaxies: formation — galaxies: high-redshift — galaxies: ISM —galaxies: star formation INTRODUCTIONUnderstanding when and how disk galaxies formedis one of the very important subjects in modern astro-physics. The cosmic star-formation rate (SFR) densityincreases from the early universe, peaks at z ∼ −
3, anddecreases by one order of magnitude to the present day(e.g., Madau & Dickinson 2014). Galaxies are thoughtto evolve drastically at z ∼ −
3, and thus observinggalaxies at this redshift range is indispensable to under-stand galaxy formation and evolution.Recent deep imagings with high angular resolutionsby
Hubble Space Telescope (HST) enable us to resolve
Corresponding author: Kaito [email protected] the structure of galaxies at z ∼ − z ∼ − z ∼ . z ∼ − z ∼ − a r X i v : . [ a s t r o - ph . GA ] J a n D e c . ( J ) V
10 kpc D e c . ( J ) composite10 kpc Figure 1.
Left: composite image of HST WFC3/IR H , J and ACS/WFC V (R, V, U band in the rest-frame, respectively)images of SXDS1 13015. The cross shows the peak position of J . Right: same as the left panel, but for the V band image.SXDS1 13015 has a UV clump at ∼ . (cid:48)(cid:48) north of the galactic center (cross mark). tional studies reveal that the clumps have stellar massof ∼ − M (cid:12) , SFR of ∼ − M (cid:12) yr − , size of1 − ∼ − yr. The clumps close to thecenter of the host galaxy show larger stellar mass, reddercolor, and older age than the clumps reside in the outerpart (e.g., F¨orster Schreiber et al. 2011; Guo et al. 2012,2018). This trend is explained by a theoretical scenariothat a gas-rich disk forms through a cold gas accretingfrom a dark matter halo and/or streaming from the out-side of the galaxy and clumps form through gravitationalinstability in the gas disk. Due to the dynamical fric-tion and/or dynamical interaction, the clumps migratetoward the center of the disk and eventually coalesceinto a young bulge (e.g., Noguchi 1999; Bournaud et al.2014).Meanwhile the inside-out growth of galaxies is alsoone aspect of observational results. For massive SFGs( M star (cid:38) M (cid:12) ), no significant increase of stellarmass is seen in the central region of galaxies since z ∼ z (cid:38)
1, while z (cid:46) M star > M (cid:12) at z ∼ −
3, Margalef-Bentabolet al. (2018) show a bulge to disk SFR ratio decreasesfrom z ∼ ∼
1, which suggests inside-out quenchingof star formation.The properties of molecular gas in SFGs at z ∼ − z ∼ − f gas = M gas / ( M gas + M star ), where M gas refers to amolecular gas mass) are high (typically, f gas ∼ .
5; e.g.,Tacconi et al. 2013; Seko et al. 2016) as compared withthose of local galaxies (typically, f gas (cid:46) . z ∼ . z ∼ − z = 1 .
45 inthe Subaru-XMM/Newton Deep Survey (SXDS; Furu-sawa et al. 2008) field. The CO observations toward thisgalaxy enable us to study molecular gas properties in anormal star-forming galaxy with a UV clump aroundat the peak of cosmic SFR density. Combining multi-transition CO data and high-resolution HST data of thegalaxy, we also study CO(5-4)/CO(2-1) flux ratio in theUV clump. Through model fittings, we compare proper-ties of molecular gas and stellar distributions, and alsoreveal the radial profile of gas mass fraction.This paper is structured as follows: in Section 2, wepresent the properties of our sample galaxy. The ALMAobservations, the data reductions, and the archival HSTdata are described in Section 3. The results and discus-sions are provided in Section 4. We report molecular gasproperties and distributions in Section 4.1, and stellardistributions in Section 4.2. In Section 4.3, we comparethe molecular gas and stellar distributions. Finally, weput the summary in Section 5. Throughout this paper,we adopt a flat cosmology with Ω M = 0 . , Ω Λ = 0 . H = 70 km s − Mpc − . At the redshift of our samplegalaxy ( z = 1 . .
45 kpc arcsec − The initial mass function (IMF) of Chabrier (2003) isadopted. Magnitudes are AB system unless otherwisenoted. SAMPLE GALAXYThe sample galaxy in this study, SXDS1 13015, is oneof the most massive main sequence galaxies at z ∼ . Ks -band selected 317 SFGs at z ∼ . ∼ . × M (cid:12) and ∼ M (cid:12) yr − , respectively.H α emission lines were detected from 71 galaxies includ-ing SXDS1 13015 and the gas metallicities were derivedbased on N2 method (Pettini & Pagel 2004).Using ALMA, Seko et al. (2016) made observationsof CO(5 −
4) and dust thermal emission toward 20SFGs selected from the sample galaxies by Yabe et al.(2012). 20 SFGs were selected to cover a wide rangeof stellar mass (4 × − × M (cid:12) ) and metallicity(12 + log(O / H) = 8 . − .
9) uniformly in the diagramsof stellar mass versus SFR and the stellar mass versus metallicity. CO(5-4) lines are detected toward 11 galax-ies and dust emissions are clearly detected from 2 galax-ies. Both the CO emission and dust thermal emissionwere detected for SXDS1 13015. Adopting metallicity-dependent CO-to-H conversion factor (Genzel et al.2012) and CO(5-4)/CO(1-0) luminosity ratio of 0.23(typical for sBzK galaxies by Daddi et al. 2015), themolecular gas mass of SXDS1 13015 was estimated tobe ∼ × M (cid:12) . Using a modified blackbody modelwith dust temperature of 30 K and dust emissivity indexof 1.5, the dust mass of SXDS1 13015 was estimated tobe ∼ × M (cid:12) .Properties of SXDS1 13015 are summarized in Ta-ble 1. Figure 1 (left) shows the composite image ofHST WFC3 F160W (hereafter, H ), F125W ( J )and ACS F606W ( V ). The galaxy seems to consistof the bulge and disk. In the disk, an arm-like feature isseen in particular northeastern side of the galaxy. Fig-ure 1 (right) shows HST V image corresponding torest-frame UV ( ∼ ∼ . (cid:48)(cid:48) Table 1.
Properties of SXDS1 13015
R.A. (J2000) a h m . s decl. (J2000) a -05 ◦ (cid:48) (cid:48)(cid:48) z COb . ± . M star ( M (cid:12) ) c,d . +0 . − . × SFR ( M (cid:12) yr − ) d ± / H) e . ± . E ( B − V ) (mag) f a Coordinate of the peak pixel of the CO(2-1) 0th momentmap. b Redshift derived from the CO(2-1) line profile. c Derived from the SED fitting with optical to mid-infrareddata. d Chabrier IMF is adopted. Difference of the adopted IMFis corrected with a factor of 0.58 (Speagle et al. 2014).The SFR is derived from the extinction-corrected UV lu-minosity density. e Derived with the N2 method. Pettini & Pagel (2004) cal-ibration is adopted. f Derived from the rest-frame UV slope. DATA SOURCES3.1.
CO Data
Observations of CO(2 −
1) toward SXDS1 13015were made with ALMA on 2016 August 24 and Septem-ber 3 during the ALMA Cycle3 (ID: 2015.1.01129.S,PI: K. Ohta). The number of 12 m antennas was 41.The length of the longest and shortest baseline was 1.8km and 15.1 m, respectively, corresponding to an an- D e c . ( J ) CO(2-1)
CO(5-4)
750 500 250 0 250 500 750velocity [km s ]0.50.00.51.01.52.02.5 C O ( - ) f l u x [ m J y ]
750 500 250 0 250 500 750velocity [km s ]210123 C O ( - ) f l u x [ m J y ] Figure 2.
Top: integrated intensity maps (0th moment maps) of CO(2-1) (left) and CO(5-4) (right) integrated over the velocityrange shown in red in the line profiles (bottom panels). The contours represent σ, σ, σ, · · · ( σ step). The black crossesshow the CO(2-1) peak position. The filled black ellipse in the bottom left corner shows the synthesized beam size. Bottom:line profiles of CO(2-1) (left) and CO(5-4) (right) in the 1. (cid:48)(cid:48) gular resolution of ∼ . (cid:48)(cid:48)
55 and a maximum recover-able scale of ∼ (cid:48)(cid:48) . The observed frequency rangewas 93 . − .
535 GHz (band 3) to detect CO(2 − ν rest = 230 .
538 GHz, ν obs = 94 .
097 GHz)with bandwidth of 937.5 MHz and spectral resolutionof 564.453 kHz, corresponding to a velocity range andresolution of 2988 km s − and 1 . − , respectively.The total on-source time was 4.7 hours. J0006-0623 andJ0238-1636 were used as the flux, bandpass, and ampli-tude calibrators. The phase calibrators were J0209-0438and J0215-0222.The details of CO(5-4) observations are described bySeko et al. (2016). The observation was conductedduring the ALMA Cycle 0 (ID: 2011.0.00648.S, PI: K.Ohta), and the angular resolution was ∼ (cid:48)(cid:48)
7. Theobserved frequency range was 222 . − .
583 GHz(band 6), the spectral resolution was 488.28 kHz ( ∼ . − ) and the on-source time was ∼
10 minutes. The raw visibility data of CO(2-1) was calibrated withthe Common Astronomy Software Applications (CASA;McMullin et al. 2007) version 4.7.2 and the observatory-provided calibration script. The raw visibility data ofCO(5-4) was calibrated with CASA version 4.2 by Sekoet al. (2016). The imagings of the ALMA data werecarried out using CASA version 5.4.0. In order to makecleaned channel maps of CO(2-1) and CO(5-4), we usedthe CASA tasks uvcontsub and tclean . Firstly, con-tinuum emission in CO(5-4) data was subtracted withthe uvcontsub . After that, we made 3 σ -clean channelmaps of CO(2-1) and CO(5-4) using the tclean withweighting of briggs ( robust = 0 .
5) and channel width of25 km s − . The synthesized beam sizes of CO(2-1) andCO(5-4) channel maps were 0 . (cid:48)(cid:48) × . (cid:48)(cid:48)
52 ( ∼ . . (cid:48)(cid:48) × . (cid:48)(cid:48)
60 ( ∼ . Table 2.
Molecular gas properties of SXDS1 13015 S CO(2 − ∆ v (Jy km s − ) 1 . ± . S CO(5 − ∆ v (Jy km s − ) 1 . ± . R
52 a . ± . L (cid:48) CO(1 − (K km s − pc ) b (5 . ± . × M mol ( M (cid:12) ) c (1 . ± . × f gas d . ± . τ depl (Gyr) e . ± . a CO(5-4)/CO(2-1) flux ratio. b We adopt CO(2-1)/CO(1-0) flux ratio of 2 (the MilkyWay like value). c Including a 36% mass contribution of helium. The metal-licity dependent CO-to- H conversion factor (Equation(7) of Genzel et al. (2015)) is adopted. d Gas mass fraction ( = M mol / ( M mol + M star ) ). e Gas depletion time ( = M mol /SF R )
25 km s − were 0 .
18 mJy beam − and 0 .
91 mJy beam − ,respectively. 3.2. HST Data
SXDS1 13015 is located in the UKIDSS UDS field,and imaged with HST ACS, WFC3/UVIS and WFC3/IRas a part of Cosmic Assembly Near-infrared Deep Ex-tragalactic Legacy Survey (CANDELS; Grogin et al.2011; Koekemoer et al. 2011) observations. We usedthe archival ACS V , WFC3 J and H data onCANDELS website . The images are drizzled and thepixel scales are 0 . (cid:48)(cid:48)
03 for V , and 0 . (cid:48)(cid:48)
06 for J and H . RESULTS AND DISCUSSIONS4.1.
Molecular Gas Properties
The integrated CO(2-1) and CO(5-4) intensitymaps (0th moment maps) and the line profiles ofSXDS1 13015 are shown in Figure 2. The symmetricdouble-peak line profile of CO(2-1) implies the galaxyhas a rotating gas disk. The peak position of the CO(5-4) image shows a slight ( ∼ . (cid:48)(cid:48)
15) offset from that ofCO(2-1) image and asymmetry is seen in the CO(5-4) line profile. These features are originated from thepresence of the UV clump as described later (Section4.1.2). Thus we adopt the peak in the CO(2-1) mapas the center of the galaxy and the central frequency ofthe observed CO(2-1) emission line (94 . ± .
008 GHz,corresponding to the redshift of 1 . ± . Total Flux and Molecular Gas Mass https://archive.stsci.edu/pub/hlsp/candels/uds/uds-tot/v1.0/ D e c . ( J ) R Figure 3.
Spatial distribution of R in SXDS1 13015. Thepixels with S / N > . in the CO(5-4) map are displayed.The contours represent , , ..., pixel values of therest UV image relative to the peak of the galaxy. The blackcross and ellipse show the CO(2-1) peak position and thesynthesized beam size, respectively. The total CO(2-1) and CO(5-4) flux of SXDS1 13015are derived by fitting an elliptical Gaussian model to therespective integrated intensity maps. We use the datain a 1. (cid:48)(cid:48) . ± .
05) Jy km s − and (1 . ± .
17) Jy km s − , re-spectively. The beam-deconvolved FWHM of the CO(2-1) source is 4 . ± . R = S CO(5 − ∆ v/S CO(2 − ∆ v ) is 1 . ± .
2, which suggeststhat the CO excitation ladder of SXDS1 13015 is simi-lar to that of the Milky Way rather than color-selectedSFGs at z ∼ − Such low COexcitations are also reported in some SFGs at z ∼ − . L (cid:48) CO(1 − )as L (cid:48) CO(1 − = 3 . × ν − − D L (1 + z ) − × R − S CO(2 − ∆ v, (1) As compared with the R -SFR surface density relations re-cently obtained by Valentino et al. (2020) and Boogaard et al.(2020), SXDS1 13015 shows a lower R at similar SFR surfacedensity among SFGs at z ∼ − where L (cid:48) CO(1 − is measured in K km s − pc , ν rest:CO(1 − is the rest frequency of the CO(1-0) emission line of115.271 GHz, D L is the luminosity distance in Mpc, R is the CO(2-1)/CO(1-0) flux ratio, and S CO(2 − ∆ v is the observed CO(2-1) flux in Jy km s − . Becausethe CO excitation ladder of the galaxy is similar tothat of the Milky Way, we assume R = 2. The cal-culated total CO(1-0) luminosity of SXDS1 13015 is(5 . ± . × K km s − pc .The molecular gas mass is derived from M mol = α CO L (cid:48) CO(1 − , (2)where α CO is the CO-to-H conversion factor in M (cid:12) (K km s − pc ) − . α CO in SFGs at z ∼ − α CO is larger in galaxies with lower metallicity (Wolfire et al.2010; Bolatto et al. 2013). As the dependence of α CO onmetallicity, we adopt the Equation (7) of Genzel et al.(2015): α CO = 4 . × − . × (12+log(O / H) − . , (3)where α CO includes a 36% mass contribution of heliumand (12 + log(O / H)) is the metallicity based on Pettini& Pagel (2004) calibration. The α CO of SXDS1 13015 iscalculated to be 2 . M (cid:12) (K km s − pc ) − . The result-ing total molecular gas mass is (1 . ± . × M (cid:12) and the gas mass fraction is f gas = 0 . ± .
08. Thegas depletion time ( τ depl = M mol / SFR) of the galaxy isestimated to be = 1 . ± . Spatial Distribution of R We made the R map by dividing the CO(5-4) 0thmoment map by the CO(2-1) map convolved to the samebeam size as CO(5-4) data. The R is calculated in pix-els with S / N > . R ∼ .
2) is signifi-cantly larger than the R averaged over SXDS1 13015and that of the Milky Way. The ratio is comparable tothe average R for sBzK galaxies at z ∼ . ∼ . (cid:48)(cid:48) ∼ −
150 km s − in the CO(2-1) velocity field (Figure 4), and this ve-locity corresponds to the peak velocity in the CO(5-4)profile (Figure 2, bottom right); the asymmetric fea-ture of the CO(5-4) profile is presumably due to thiscomponent. We show the 0th moment maps integratedover the velocity range of −
250 km s − to 0 km s − in D e c . ( J ) v e l o c i t y [ k m s ] Figure 4.
Velocity field (1st moment map) of SXDS1 13015.The contours show the rest UV distribution (same as Figure3). The black cross and solid line show the CO(2-1) peakposition and the major axis of the galaxy, respectively. Thefilled black ellipse in the bottom left corner shows the synthe-sized beam size.
Figure 5. The R map in this velocity range also peaksat the position of the UV clump. In the UV clump, itis expected that the density and/or temperature of themolecular gas are higher than those in other regions.This result, the high R in the UV clump, is qualita-tively consistent with the CO excitation ladder modeledby a large velocity gradient analysis of the hydrodynam-ically simulated high- z clumpy galaxy (Bournaud et al.2015).It is worth noting that the spatial offset and theasymmetric profile of CO(5-4) are not results of noise.We performed simulations of CO(5-4) observation withCASA task simobserve assuming a gaussian distribu-tion mimicking the CO(5-4) distribution and line pro-file. Consequently, we found that the 1- σ uncertaintyof the peak position of CO(5-4) is ∼ .
04 arcsec. heobserved offset between the peak positions of CO(5-4) and CO(2-1) is ∼ .
15 arcsec which corresponds to ∼ . σ . We also found that the 1- σ uncertainty ofthe peak velocity is ∼
30 km s − . The observed peakCO(5-4) velocity and the central CO(2-1) velocity is ∼
150 km s − which corresponds to ∼ σ . Thus theoffset and the peak velocity difference are consideredto be real. It should be also noted here, we made as-trometric corrections to the HST images with the off-set shown below. We assume the peak of J imagecoincides with the peak of the CO(2-1) 0th momentmap, because the CO(2-1) shows a good symmetry inthe line profile and S/N of the image is high. The off- D e c . ( J ) CO(2-1)
CO(5-4) D e c . ( J ) R Figure 5.
Top left: CO(2-1) 0th moment map integrated over the velocity range of −
250 km s − to − . The contoursshow the rest UV distribution (same as Figure 3). The black cross and the ellipse show the CO(2-1) peak position in the totalintegrated map and the synthesized beam size, respectively. Top right: same as top left panel, but for CO(5-4). Bottom left:same as Figure 3, but for the velocity range of −
250 km s − to − . set is (∆R . A ., ∆Dec . ) = (+0 . (cid:48)(cid:48) , − . (cid:48)(cid:48) z ∼ .
5, but no CO(5-4) emission from the UV clumpswas detected. This is presumably due to the low SFRof the UV clumps ( ∼ − M (cid:12) yr − ). The sensitivityof their observation was not deep enough to detect theCO(5-4) emissions from the UV clumps. The SFR ofthe UV clump in our sample galaxy is estimated to be ∼ M (cid:12) yr − from UV luminosity, and is larger thanthose of clumps in Cibinel et al. (2017). By assuminga depletion time of 0 . J -ladder, the contribution of the UV clump to the totalCO(5-4) flux is expected to be ∼ −
25% in our case.4.1.3.
Fitting a Rotating Disk Model
Table 3.
Best-fit parameters derived by
GalPaK flux (Jy km s − ) 1 . ± . r / (kpc) a . ± . . ± . . ± . r t (kpc) b . ± . V max (km s − ) c ± Uncertainties are estimated based on simulations (seetext). a Half-light radius. b Turnover radius. c Inclination-corrected maximum rotational velocity.
The velocity field of CO(2-1) (Figure 4) shows a clearvelocity gradient along the major axis, indicating thatthe galaxy has a rotating gas disk, which is also im-plied by the double-peak CO(2-1) line profile. In or-der to eliminate the effect of beam smearing and derivethe intrinsic molecular gas distribution and kinematicproperties of the galaxy, we fit a rotating disk model D e c . ( J ) observe model D e c . ( J ) residual C O ( - ) f l u x [ m J y ] observemodel750 500 250 0 250 500 750velocity [kms ]2.50.02.5 r e s i d u a l [] Figure 6.
Top left: observed integrated intensity map of CO(2-1). The filled black ellipse in the bottom left corner showsthe synthesized beam size. Top right and bottom left: Same as top left, but for best-fit model through
GalPaK and residual( observed − model ), respectively. The contours in the residual map represent − σ, − σ, σ and σ . Bottom right: observed andmodeled line profiles (red and blue solid curve, respectively). The residual normalized with σ noise level is shown in lower part. by using GalPaK (Galaxy Parameters and Kinemat-ics; Bouch´e et al. 2015). GalPaK is a program codethat estimates morphological parameters (e.g., size andinclination) and kinematic parameters (e.g., maximumrotational velocity and velocity dispersion) of the galaxyby fitting a disk model convolved with two-dimensionalPSF and one-dimensional line spread function (LSF) toa given data cube. We fit a disk model whose radialflux profile is exponential (S´ersic index n = 1) and ra-dial rotational velocity profile is hyperbolic tangent tothe CO(2-1) channel map. PSF and LSF correspond tothe synthesized beam (0 . (cid:48)(cid:48) × . (cid:48)(cid:48)
53, PA = 79 ◦ ) andthe channel bin (velocity width is 25 km s − ), respec-tively. The fitting parameters are central position, cen-tral frequency, total flux, half-light radius ( r / ), incli-nation, position angle, turnover radius ( r t ), maximumrotational velocity ( V max ), and velocity dispersion ( σ )which is assumed to be constant over the disk. Thebest-fit models as well as input image are shown in Fig-ure 6 (upper panels). The residual (observed − model) integrated intensity map and line profile are also shownin Figure 6 (bottom panels). The residuals demonstratethat the CO(2-1) distribution of the galaxy is well rep-resented by the rotating disk model. The best-fit pa-rameters are listed in Table 3. In order to estimate theuncertainties on the best-fit parameters, we created 100cleaned cubes by simulating observations of the best-fit disk model with simobserve by changing the noiseseeds randomly. Disk models are fitted to the simulatedcubes by GalPaK , and the means and the standarddeviations of the disk model parameters are derived.We adopt the square root of the sum of the square ofsystematic error and standard deviation as the uncer-tainty for each best-fit parameter derived by GalPaK .The best-fit flux (1 . ± .
05 Jy km s − ) is consistentwith that derived from the two-dimensional Gaussianfitting to the integrated intensity map. The best-fit r / = 2 . ± .
12 kpc is also consistent with the HWHM
Table 4.
Results of fitting S´ersic profile models to J and H images filter J H n a . ± .
19 1 . ± . R / (kpc) a . ± .
15 3 . ± . b/a a . ± .
01 0 . ± . a . ± . . ± . R / (kpc) b . ± .
23 0 . ± . R / (kpc) b . ± .
33 4 . ± . L bulge /L disk b . +0 . − . . +0 . − . Uncertainties are estimated by changing sky level within its error, stars to construct the PSFsand the size of the fitting region. a Results of single component fitting. S´ersic index, half-light radius, axis ratio, and position angle,respectively. b Results of two components (bulge ( n = 4 ) and disk ( n = 1 )) fitting. [ m a g a r c s e c ] bulge disksinglebulge+diskobserved( J ) 0.0 0.5 1.0 1.5major axis radius [arcsec]0.0 2.5 5.0 7.5 10.0 12.5 15.0major axis radius [kpc]2021222324252627 [ m a g a r c s e c ] bulge disksinglebulge+diskobserved( H ) Figure 7.
Left: J radial profiles of SXDS1 13015. µ is the surface brightness. Gray circles with errorbars show theobserved radial profile. The best-fit single component model and two components (bulge + disk) model are shown with blue andorange solid line, respectively. Dashed orange lines show the bulge and disk components of the best-fit two components model.Right: same as the left panel, but for H . of the 2D Gaussian fitting ( ∼ . The best-fit V max and σ are 350 ±
30 km s − and 2 . ± . − , re-spectively. The best-fit value for σ is very much smalland has a very large uncertainty ( σ is derived to be ∼
40 km s − from some simulated cubes). The value isconsidered not to be reliable. In fact, the observed lineprofile is well reproduced without changing other out-put parameters such as r / , even if we fit a disk modelwith a fixed σ ∼
50 km s − (typical value found in H α observations for z < Generally, the HWHM of a 2D Gaussian corresponds to half-light radius. by beam smearing effect, and it is hard to constrain theintrinsic velocity dispersion well.4.2.
Stellar Distribution
In order to derive properties of the stellar distribu-tion, we fit S´ersic profile models to HST WFC3/IR J and H band images using GALFIT (Peng et al. 2002).GALFIT is a program code that fits given surface bright-ness model convolved with PSF to an observed galaxy.Four(Six) bright but unsaturated stars in the J ( H )image of the CANDELS-UDS field are stacked to con-struct the PSFs for the fittings. The ∼ . (cid:48)(cid:48) × . (cid:48)(cid:48) l o g ( s t a r [ M k p c ]) SXDS1_13015 ( J ) z = 0.06 z = 1.0 1.5 z = 1.5 2.0 Figure 8.
The red line shows the stellar mass surface densityprofile of SXDS1 13015 with the uncertainty derived from theerror of the stellar mass and galaxy parameters of GALFIT.A constant mass-to-luminosity ratio is assumed. The green,blue, and black dashed lines show the median profiles of galax-ies at z = 1 . − . , . − . , and . , respectively (Patelet al. 2013). The galaxies are selected at a constant cumu-lative number density of . × − Mpc − for different red-shifts (corresponding to the stellar mass of ∼ . × M (cid:12) for . < z < . ). We show the standard deviation of eachmedian profile (see text). The gray shaded region shows theHWHM of PSFs for our fitting. The fitting parameters are central position, total mag-nitude, half-light radius ( R / ), S´ersic index ( n ), axialratio, and position angle. We also fit another surfacebrightness model with two-component S´ersic profiles. Inthis fitting, S´ersic indices are fixed to be n = 4 for bulgeand 1 for disk, because the galaxy has a rotating gas diskand seems to consist of the bulge and disk in the HSTcomposite image (see Figure 1). The fitting parametersare total magnitude of each component, half-light radiusof each component ( R / , R / ) and axial ratioof the bulge. Central position of each component, posi-tion angle of each component, and axial ratio of the diskare fixed to the values derived from the single compo-nent fitting. The derived best-fit parameters are listedin Table 4. The observed and fitted radial profiles areshown in Figure 7.Assuming a constant mass-to-luminosity ratio, wederived the stellar mass surface density profile ofSXDS1 13015 from J distribution modeled with thesingle component model. In Figure 8, the derived profileof SXDS1 13015 is plotted with the median profiles ofgalaxies at different redshifts by Patel et al. (2013). Thesample galaxies of Patel et al. (2013) are selected at a constant cumulative number density of 1 . × − Mpc − for different redshifts (corresponding to the stellar massof ∼ . × M (cid:12) for 1 . < z < . . < z < . ∼ z ∼ . − . z ∼ . − . z ∼ . r (cid:46) ∼ . ∼ . Molecular Gas and Stellar Distribution inSXDS1 13015
Modeled Gas and Stellar Distributions
By using the modeled molecular gas distribution de-rived by GalPaK and the stellar distribution derivedby GALFIT, we obtain intrinsic radial surface distribu-tions of the molecular gas as well as the stellar compo-nent in SXDS1 13015 (Figure 9). The uncertainty onthe molecular gas surface distribution is estimated withradial profiles derived from GalPaK fitting to the 100simulated cubes (see Section 4.1.3). The uncertainty onthe stellar surface distribution is derived from uncer-tainties on the total stellar mass and galaxy parametersof GALFIT. We also derive gas mass fraction ( f gas ) ateach radius. The uncertainty on f gas is derived fromthe uncertainties on the molecular gas and stellar radialsurface distributions. f gas is higher than galactic totalvalue ( ∼ .
55) at r (cid:46) ∼ . ∼ . ∼ × r / . The molecular gas is distributedin more inner region of the galaxy than stars and seemsto associate with the bulge rather than the stellar disk,which is also indicated by the best-fit half-light radii.By using the best-fit kinematic parameters of themolecular gas disk, dynamical masses ( M dyn ( r ) ∼ rv /G ) at r = r / , r / , r / ( r / is half-lightradius of molecular gas disk) are estimated to be1 Figure 9.
Left: Modeled (intrinsic) molecular gas and stellar ( J ) radial distributions of SXDS1 13015. Green and blue lineshows the molecular gas and stellar (one component) surface density profile, respectively. Shaded regions show the uncertainties.The gas mass fraction ( f gas = M mol / ( M mol + M star ) ) is also plotted with red line, and the shaded region shows the uncertaintyof f gas . Right: same as the left panel, but for H . ∼ (0 . , . , . × M (cid:12) . We also derive the molec-ular gas and stellar masses within these radii assumingthe inner mass ( m ( < r )) is proportional to the luminos-ity within the radius ( L ( < r )) in J image: m ( < r ) = L ( < r ) L total M total . (4)The molecular gas and stellar masses within r = r / , r / , r / are m mol ∼ (0 . , . , . × M (cid:12) and m star ∼ (0 . , . , . × M (cid:12) , respectively.Baryon fractions ( f baryon = ( m mol + m star ) / ( M dyn + m mol + m star )) within those radii are estimated to be ∼ . , . , .
53; within r = 3 r / ∼ CO and Optical Half-light Radii
Figure 10 (left) shows the location of SXDS1 13015 inoptical versus CO half-light radius plane compared toother field and cluster galaxies at z ∼ − .
6. The fieldsample includes 12 main sequence galaxies at z ∼ − . n = 1 S´ersic profile to image planesor circular Gaussians to uv planes, and by fitting sin-gle S´ersic profile to HST-WFC3 I band data usingGALFIT, respectively. The typical uncertainties of bothCO and optical half-light radius is ∼ z ∼ . H band data using GALFIT, respectively. The typ-ical uncertainty of CO half-light radius is ∼ V band of galaxies at similar redshift toSXDS1 13015 show that SFGs with similar stellar massshow half-light radius of a few to a few tens kpc andS´ersic index of n ∼ . − n and half-light radius of SXDS1 13015 de-rived from the single component fitting are within therange of those obtained in other main sequence galax-ies. The CO and optical half-light radii of SXDS1 13015are, however, relatively small as compared with othermain sequence galaxies, but the CO-to-optical half-lightradius ratio of SXDS1 13015 ( R / , CO /R / , F125W =0 . ± . R / , CO /R / , F160W = 0 . ± .
04) is typicalamong these samples. SUMMARYWe presented the results of sub-arcsecond ALMA CO(2-1) and CO(5-4) observations toward a mas-sive main sequence galaxy at z = 1 .
45 (SXDS1 13015).These observations enabled us to study molecular gasproperties, its distribution in the galaxy, and CO(5-4)/CO(2-1) flux ratio in a UV clump detected in therest-frame UV HST image. By fitting a rotating gasdisk model to the CO(2-1) data, we derived molecular2 M star [ M ]1.01.52.02.53.0 S F R [ M y r ] field (Tacconi+13, Daddi+10)cluster (Noble+19)SXDS1_130150 2 4 6 8 10 12optical R [kpc]024681012 C O R / [ k p c ] field (Tacconi+13, Daddi+10)cluster (Noble+19)SXDS1_13015 (f125w)SXDS1_13015 (f160w) Figure 10.
Left: Distribution of the galaxies in optical versus CO half-light radius plane. SXDS1 13015, field galaxies, andcluster galaxies are shown in red and orange squares, blue circles, and green circles, respectively. For SXDS1 13015, the resultsof the single component fittings to J and H images are plotted in red and orange, respectively. The field galaxies are takenfrom Tacconi et al. (2013) and Daddi et al. (2010) (classified “Disk(A)” in Tacconi et al. (2013)). The cluster galaxies are takenfrom Noble et al. (2019) (with high S/N and clear velocity gradient in CO(2-1)). The dashed line shows optical equals CO. Right:stellar mass versus SFR. SXDS1 13015, field galaxies, and cluster galaxies are shown in red, blue, and green, respectively. Themain sequence at z = 1 . (Speagle et al. 2014) is shown with solid line together with the scatter (dashed lines). The differenceof adopted IMF is corrected with the factor by Speagle et al. (2014). gas distribution and kinematics. Combining with theHST images, we compared the properties of moleculargas and stellar distributions. The results are as follows:i) CO(2-1) and CO(5-4) emission lines are clearly de-tected from the galaxy (Figure 2) and the symmetric anddouble-peak line profile of CO(2-1) implies the presenceof the molecular gas disk in the galaxy. R of the galaxyis 1 . ± .
2, which suggests the CO excitation ladder ofthe galaxy is similar to that of the Milky Way ratherthan sBzK galaxies at z ∼ .
5. The molecular gas massof the galaxy is 1 . × M (cid:12) , adopting R = 1 (theMilky Way value) and metallicity dependent CO-to-H conversion factor. The gas mass fraction and depletiontime are 0 . ± .
08 and 1 . ± . R map (Figure 3), the peak value of R is ∼ .
2, comparable to that of the average of sBzK galax-ies at z ∼ . R peaks at the position of the UVclump. These results are similar to a result of a hydro-dynamic simulation of a clumpy galaxy (Bournaud et al.2015). In the UV clump, the gas density and/or tem-perature are/is higher than those in the other galacticregions.iii) By using GalPaK , the molecular gas distributionof the galaxy traced by CO(2-1) is well represented bythe rotating disk model (Figure 6). The half-light radiusof the modeled gas disk is 2 . ∼ . ∼ . ∼ × r / , suggesting that the galaxy is forming its bulge.We thank the anonymous referee for useful commentsand suggestions, which improve the paper. We are grate-ful to K. Nakanishi, F. Egusa, K. Saigo, and the staffat the ALMA Regional Center for their help in datareduction. We also thank K. Tadaki for his critical com-ment on our earlier study with this galaxy. F.M. is sup-ported by Research Fellowship for Young Scientists fromthe Japan Society of the Promotion of Science (JSPS).K.O. is supported by JSPS KAKENHI Grant NumberJP19K03928. This paper makes use of the followingALMA data: ADS/JAO.ALMA Software:
GALFIT (Peng et al. 2002), GalPaK3D(Bouch´e et al. 2015), CASA (v4.7.2; McMullin et al. 2007)3REFERENCES
Barro, G., Kriek, M., P´erez-Gonz´alez, P. G., et al. 2016,ApJL, 827, L32, doi: 10.3847/2041-8205/827/2/L32Bolatto, A. D., Wolfire, M., & Leroy, A. K. 2013, ARA&A,51, 207, doi: 10.1146/annurev-astro-082812-140944Boogaard, L. A., van der Werf, P., Weiß, A., et al. 2020,arXiv e-prints, arXiv:2009.04348.https://arxiv.org/abs/2009.04348Bouch´e, N., Carfantan, H., Schroetter, I., Michel-Dansac,L., & Contini, T. 2015, AJ, 150, 92,doi: 10.1088/0004-6256/150/3/92Bournaud, F., Daddi, E., Weiß, A., et al. 2015, A&A, 575,A56, doi: 10.1051/0004-6361/201425078Bournaud, F., Perret, V., Renaud, F., et al. 2014, ApJ, 780,57, doi: 10.1088/0004-637X/780/1/57Bruzual, G., & Charlot, S. 2003, MNRAS, 344, 1000,doi: 10.1046/j.1365-8711.2003.06897.xCalzetti, D., Armus, L., Bohlin, R. C., et al. 2000, ApJ,533, 682, doi: 10.1086/308692Cameron, E., Carollo, C. M., Oesch, P. A., et al. 2011, ApJ,743, 146, doi: 10.1088/0004-637X/743/2/146Carilli, C. L., & Walter, F. 2013, ARA&A, 51, 105,doi: 10.1146/annurev-astro-082812-140953Chabrier, G. 2003, PASP, 115, 763, doi: 10.1086/376392Cibinel, A., Daddi, E., Bournaud, F., et al. 2017, MNRAS,469, 4683, doi: 10.1093/mnras/stx1112Daddi, E., Bournaud, F., Walter, F., et al. 2010, ApJ, 713,686, doi: 10.1088/0004-637X/713/1/686Daddi, E., Dannerbauer, H., Liu, D., et al. 2015, A&A, 577,A46, doi: 10.1051/0004-6361/201425043Decarli, R., Walter, F., Aravena, M., et al. 2016, ApJ, 833,70, doi: 10.3847/1538-4357/833/1/70Decarli, R., Walter, F., G´onzalez-L´opez, J., et al. 2019,ApJ, 882, 138, doi: 10.3847/1538-4357/ab30feDunlop, J. S., McLure, R. J., Biggs, A. D., et al. 2017,MNRAS, 466, 861, doi: 10.1093/mnras/stw3088Elmegreen, D. M., Elmegreen, B. G., Ravindranath, S., &Coe, D. A. 2007, ApJ, 658, 763, doi: 10.1086/511667F¨orster Schreiber, N. M., Shapley, A. E., Genzel, R., et al.2011, ApJ, 739, 45, doi: 10.1088/0004-637X/739/1/45F¨orster Schreiber, N. M., Renzini, A., Mancini, C., et al.2018, ApJS, 238, 21, doi: 10.3847/1538-4365/aadd49Furusawa, H., Kosugi, G., Akiyama, M., et al. 2008, ApJS,176, 1, doi: 10.1086/527321Genzel, R., Tacconi, L. J., Combes, F., et al. 2012, ApJ,746, 69, doi: 10.1088/0004-637X/746/1/69Genzel, R., Tacconi, L. J., Lutz, D., et al. 2015, ApJ, 800,20, doi: 10.1088/0004-637X/800/1/20Grogin, N. A., Kocevski, D. D., Faber, S. M., et al. 2011,ApJS, 197, 35, doi: 10.1088/0067-0049/197/2/35 Guo, Y., Giavalisco, M., Ferguson, H. C., Cassata, P., &Koekemoer, A. M. 2012, ApJ, 757, 120,doi: 10.1088/0004-637X/757/2/120Guo, Y., Ferguson, H. C., Bell, E. F., et al. 2015, ApJ, 800,39, doi: 10.1088/0004-637X/800/1/39Guo, Y., Rafelski, M., Bell, E. F., et al. 2018, ApJ, 853,108, doi: 10.3847/1538-4357/aaa018Hodge, J. A., & da Cunha, E. 2020, arXiv e-prints,arXiv:2004.00934. https://arxiv.org/abs/2004.00934Kennicutt, Robert C., J. 1998, ARA&A, 36, 189,doi: 10.1146/annurev.astro.36.1.189Koekemoer, A. M., Faber, S. M., Ferguson, H. C., et al.2011, ApJS, 197, 36, doi: 10.1088/0067-0049/197/2/36Lawrence, A., Warren, S. J., Almaini, O., et al. 2007,MNRAS, 379, 1599,doi: 10.1111/j.1365-2966.2007.12040.xLilly, S., Schade, D., Ellis, R., et al. 1998, ApJ, 500, 75,doi: 10.1086/305713Madau, P., & Dickinson, M. 2014, ARA&A, 52, 415,doi: 10.1146/annurev-astro-081811-125615Maeda, F., Ohta, K., & Seko, A. 2017, ApJ, 835, 120,doi: 10.3847/1538-4357/835/2/120Margalef-Bentabol, B., Conselice, C. J., Mortlock, A., et al.2018, MNRAS, 473, 5370, doi: 10.1093/mnras/stx2633McMullin, J. P., Waters, B., Schiebel, D., Young, W., &Golap, K. 2007, in Astronomical Society of the PacificConference Series, Vol. 376, Astronomical Data AnalysisSoftware and Systems XVI, ed. R. A. Shaw, F. Hill, &D. J. Bell, 127Mortlock, A., Conselice, C. J., Hartley, W. G., et al. 2013,MNRAS, 433, 1185, doi: 10.1093/mnras/stt793Mowla, L. A., van Dokkum, P., Brammer, G. B., et al.2019, ApJ, 880, 57, doi: 10.3847/1538-4357/ab290aNoble, A. G., Muzzin, A., McDonald, M., et al. 2019, ApJ,870, 56, doi: 10.3847/1538-4357/aaf1c6Noguchi, M. 1999, ApJ, 514, 77, doi: 10.1086/306932Patel, S. G., van Dokkum, P. G., Franx, M., et al. 2013,ApJ, 766, 15, doi: 10.1088/0004-637X/766/1/15Peng, C. Y., Ho, L. C., Impey, C. D., & Rix, H.-W. 2002,AJ, 124, 266, doi: 10.1086/340952Pettini, M., & Pagel, B. E. J. 2004, MNRAS, 348, L59,doi: 10.1111/j.1365-2966.2004.07591.xRiechers, D. A., Pavesi, R., Sharon, C. E., et al. 2019, ApJ,872, 7, doi: 10.3847/1538-4357/aafc27Rujopakarn, W., Dunlop, J. S., Rieke, G. H., et al. 2016,ApJ, 833, 12, doi: 10.3847/0004-637X/833/1/12Salpeter, E. E. 1955, ApJ, 121, 161, doi: 10.1086/145971Seko, A., Ohta, K., Yabe, K., et al. 2016, ApJ, 819, 82,doi: 10.3847/0004-637X/819/1/824