Molecular gas distribution perpendicular to the Galactic plane
Yang Su, Ji Yang, Qing-Zeng Yan, Shaobo Zhang, Hongchi Wang, Yan Sun, Zhiwei Chen, Chen Wang, Xin Zhou, Xuepeng Chen, Zhibo Jiang, Min Wang
aa r X i v : . [ a s t r o - ph . GA ] F e b Draft version February 23, 2021
Typeset using L A TEX preprint2 style in AASTeX63
Molecular gas distribution perpendicular to the Galactic plane
Yang Su, Ji Yang,
1, 2
Qing-Zeng Yan, Shaobo Zhang, Hongchi Wang,
1, 2
Yan Sun, Zhiwei Chen, Chen Wang, Xin Zhou, Xuepeng Chen,
1, 2
Zhibo Jiang, and Min Wang Purple Mountain Observatory and Key Laboratory of Radio Astronomy, Chinese Academy of Sciences, Nanjing210023, China School of Astronomy and Space Science, University of Science and Technology of China, 96 Jinzhai Road, Hefei230026, China
ABSTRACTWe use the ∼
370 square degrees data from the MWISP CO survey to study thevertical distribution of the molecular clouds (MCs) toward the tangent points in theregion of l = [+16 ◦ , +52 ◦ ] and | b | < . ◦
1. We find that the molecular disk consists of twocomponents with the layer thickness (FWHM) of ∼
85 pc and ∼
280 pc, respectively. Inthe inner Galaxy, the molecular mass in the thin disk is dominant, while the molecularmass traced by the discrete MCs with weak CO emission in the thick disk is probably < ∼
10% of the whole molecular disk. For the CO gas in the thick disk, we identified 1055high- z MCs that are > ∼
100 pc from the Galactic plane. However, only a few samples(i.e., 32 MCs or 3%) are located in the | z | > ∼
360 pc region. Typically, the discreteMCs of the thick disk population have a median peak temperature of 2.1 K, a medianvelocity dispersion of 0.8 km s − , and a median effective radius of 2.5 pc. Assuminga constant value of X CO = 2 × cm − (K km s − ) − , the median surface density ofthese MCs is 6 . M ⊙ pc − , indicating very faint CO emission for the high- z gas. Thecloud-cloud velocity dispersion is 4.9 ± . − and a linear variation with a slope of ∼ − . − kpc − is obtained in the region of R GC =2.2–6.4 kpc. Assuming that theseclouds are supported by their turbulent motions against the gravitational pull of thedisk, a model of ρ ( R ) = 1 . M ⊙ pc − e − R . can be used to describe the distributionof the total mass density in the Galactic midplane. Keywords:
Interstellar medium (847); Molecular clouds (1072); Surveys (1671); MilkyWay disk (1050); Interstellar dynamics (839); Stellar-interstellar interactions(1576) INTRODUCTIONCO is a good tracer of the molecular gas in theinterstellar medium (ISM). Many CO surveys(e.g., see Heyer & Dame 2015; Schuller et al.
Corresponding author: Yang [email protected] R GC > ∼
10 kpc in Su et al. 2016) often dis-plays the flaring and warping structures.Additionally, the new CO data reveal thatthe considerable molecular gas is located inthe region away from the b = 0 ◦ plane inthe inner Galaxy because of the larger lat-itude coverage of b = [ − . ◦
1, 5 . ◦
1] in theMWISP survey (Su et al. 2019). The simi-lar thick molecular disk was investigated bothin the Milky Way (Dame & Thaddeus 1994;Malhotra 1994a) and other galaxies (e.g., NGC891 by Garcia-Burillo et al. 1992 and M51 byPety et al. 2013). Following the study byDame & Thaddeus (1994), we have confirmedthat the CO disk in the inner Galactic planeis composed of two components, a well-knowthin disk with a full width at half maximum(FWHM) of 88.5 pc and an additional faintthick CO disk with a FWHM of 276.8 pc. How-ever, the properties and distributions of the molecular gas in the Galactic thick disk havenot been studied in detail due to lacking highquality large-scale data and large MC samples.In this paper, we focus on the vertical distribu-tion of the molecular gas near the tangent pointsby using the accumulated CO data from theMWISP. Benefiting from the good angular res-olution, excellent sensitivity, and high dynamicrange of the MWISP survey, we identified over1000 MCs far from the Galactic plane to tracethe thick CO disk in the region of l = 16 ◦ –52 ◦ and | b | < ∼ . ◦
1. The properties, distributions, andorigins of these MCs far from the Galactic plane(hereafter the high- z MCs) are investigated ac-cordingly.The paper is organized as follows. In Sec-tion 2, we briefly describe the CO observationsand data processing. Section 3 shows the re-sults and discussions about the CO distribu-tion perpendicular to the Galactic disk and theMCs far from the midplane. The distributionof the total midplane mass density as a func-tion of the Galactocentric distance is also inves-tigated in the section based on the estimationsof the cloud-cloud velocity dispersion and thescale height of the CO gas. Finally, we summa-rize our results in Section 4. CO DATAWe employed the ∼
370 deg CO data (i.e., l = 16 ◦ –52 ◦ and | b | < ∼ . ◦
1) from the MWISPproject (see the details in Su et al. 2019) tostudy the vertical distribution of the moleculargas. Briefly, the CO, CO, and C O ( J =1–0) lines were simultaneously observed with thefull-sampling On-The-Fly mapping (OTF; seeSun et al. 2018) by using the nine-beam Super-conducting Spectroscopic Array Receiver sys-tem (Shan et al. 2012). The spatial and spec-tral resolutions of the CO data are ∼ ′′ and ∼ − , respectively. The quality of theCO data is good and the first-order (or lin-ear) baseline was fitted for all spectra. Af-ter removing the bad channels and abnormal ertical Distribution of MCs ′′ have typical rms noiselevels of ∼ CO at a channel widthof 0.16 km s − and ∼ CO (C O) at0.17 km s − , respectively. RESULTS AND DISCUSSIONS3.1.
Terminal Velocities from the MWISP COData
Figure 1 displays the schematic view of themolecular gas discussed in this paper. The redbelt with a length of ∼ l = 16 ◦ and l = 52 ◦ . The blue circle indicates the approx-imate radius of the end of the bar (i.e., the 3-kpc ring structure). The gas emission inside ofthe circle drops sharply. Assuming the circularmotion of MCs in the Milky Way, we can easilyobtain the distances ( d = R × cos l ) and Galac-tocentric distances ( R GC = R × sin l ) for theMCs at the terminal velocity. Here the value ofthe Sun’s distance to the Galactic center, R ,is taken to be 8.15 kpc (see the recent work byReid et al. 2019).The CO emission near the tangent points ismainly concentrated in the region of | b | < ∼ . ◦ l = 16 ◦ − ◦ . Each sub-region is centered atb=0 ◦ and has an area of 3 deg , i.e., ∆ l = 1 ◦ and b = [ − . ◦
5, 1 . ◦ V tan ( l ),for each sampled Galactic longitude. At thetangent point, the typical value of the averagepeak CO temperature of each sub-region is se-lected as T peak12 > ∼ l = 22 ◦ − ◦ and T peak12 > ∼ l = 16 ◦ − ◦ . That is, the smaller features (and thus the MCs with veryweak CO emission) are ignored since we onlyfocus on the ensemble of the large-scale molec-ular gas near the tangent points in each sub-region. These smaller features with somewhatlarger V LSR are probably due to local variationsof the environment (i.e., the abnormal veloci-ties from star-forming activities, e.g., H ii re-gions, star winds, supernova remnants, etc) andnoncircular motions from other effects near thetangent points. For blended and confused COemission, the combined CO line is also used todefine the CO peak emission at the most posi-tive velocity (i.e., V tan ( l ) at T peak13 > ∼ l = [16 ◦ , ◦ ] (see the red linein Figure 2). We find that our result is consis-tent with previous CO studies (e.g., see Clemens1985; Malhotra 1994b).For further comparisons of the tangent-point features at the most positive veloc-ity, we also overlaid the terminal velocitycurves from the H i observations (the goldlines, McClure-Griffiths & Dickey 2016) andthe trigonometric parallaxes of high-mass star-forming regions (the purple line, see the A5model from Reid et al. 2019) on the CO LVmap, respectively. Generally, for the discussedregion the l – V tan trend is very similar betweenour CO result and the H i result, although ourCO terminal velocity is systematic smaller (sev-eral km s − ) than that of the H i observations.Actually, we find that the H i terminal-velocitycurve is nearly around the outer layer of theCO emission at the most extreme velocity (seethe thick golden line in Figure 2). The differ-ence between the CO result and the H i result isprobably due to (1) the different properties ofthe two tracers, and (2) the different methodsused in studies.On the other hand, models based on differenttracers also display somewhat difference in cal-culating the terminal velocity at different Galac-tic longitudes (e.g., the thin golden line vs. thepurple line in Figure 2). Further investigationsare helpful to clarify these issues by the com-binations of the H i and CO data in a largerlongitude range. In this paper, we tentativelyuse the MWISP CO observations to trace theterminal velocity (see the red line in Figure 2)for the subsequent analysis. As discussed below,the slight variation of the terminal velocity doesnot change our analysis and result considerably.3.2. Thin and Thick Molecular Disks
Using the determined terminal velocity curvein Section 3.1, we can investigate the verticaldistribution of the molecular gas at the tangentpoints. We made the integrated CO emissionin the velocity range of V ( l ) > ∼ V tan ( l ) − − to trace the molecular gas at the most positivevelocity (Figure 3). Some large-scale structuresnear the tangent points are labeled on the map(e.g., refer to Bland-Hawthorn & Gerhard 2016;Reid et al. 2016). The value of 7 km s − wasadopted to take into account the cloud-cloud ve-locity dispersion of the molecular gas near thetangent points (e.g., see Malhotra 1994b). Ac-tually, this value is roughly consistent with themaximum cloud-cloud velocity dispersion mea-sured from the tangent-point MCs based on theMWISP CO observations (Section 3.4).Based on the CO emission near the mostpositive velocity (Figure 3), we find that molec-ular gas is mainly concentrated around theplane in the inner Galaxy (e.g., | b | < ∼ ◦ or | z | < ∼
100 pc). The whole region from l =[+16 ◦ , +52 ◦ ] is divided into six ranges of lon-gitude to investigate the vertical distribution of the molecular gas. For a certain region, COemission at the same distance from the Galac-tic plane of b = 0 ◦ (i.e., z = R GC cos( l )tan( b )) isintegrated assuming the gas is located at thetangent points. Along the direction perpen-dicular to the Galactic plane, the total inten-sity of the CO emission with a bin of 5 pc isthen fitted by using a Gaussian function, i.e., I ( z ) /I peak =exp( − ( z − z ) σ ). Here, z and σ z (i.e.,the thickness FWHM=2.355 σ z ) in units of pcis the mean value and the standard deviationof the vertical distribution of the CO emission,respectively. Note that the CO gas outsidethe | z | > ∼ z emis-sion is very weak in the total CO intensity nearthe tangent points.The best fit of the vertical distribution of the CO emission is shown in Figure 4, while theresult of the CO emission is also shown forcomparison. Obviously, the z values from the CO emission agree well with those from the CO emission, indicating that the moleculargas traced by the optically thick and thin linesis concentrated toward the midplane. We findthat the mean thickness of the CO disk isroughly 94 pc between l = 22 ◦ and l = 52 ◦ ,which is larger than the value of ∼
60 pc in theregion of l = 16 ◦ − ◦ . On the whole, thenarrow Gaussian component well represents thethin molecular gas disk that is dominated by theenhanced CO emission near the Galactic mid-plane.Meanwhile, there are amounts of CO gas lo-cated far from the Galactic plane (Figure 3). Toinvestigate the distribution of the gas far fromthe plane, we use the mean intensity in each5 pc bin to reveal the weak CO emission. Thatis, we use the mean intensity in an effective area(i.e., I total / p N pixel for pixels of I CO > ∼ I rms ) toinvestigate the distribution of the CO gas inthe disk. Obviously, this processing nearly doesnot affect the distribution of the gas in the thin ertical Distribution of MCs z region becauseonly pixels with I CO > ∼ I rms are accounted for.By adopting the known values of z and σ z forthe thin CO disk (i.e., the narrow Gaussiancomponent from Figure 4), we can easily fit an-other broad Gaussian component for the thickCO disk (Figure 5).As a result, a mean FWHM of ∼
256 pc (ex-cluding the region of l = 16 ◦ − ◦ ) can well de-scribe the thick molecular gas disk. It is worthmentioning that the gas distribution of the thickdisk is not symmetric with respect to the thinCO disk in some regions. Then the fitted z val-ues of the thick disk are not always consistentwith those from the thin CO disk (e.g., see theenhanced CO emission at z < ∼ −
100 pc in thelongitude range of 22 ◦ − ◦ in Figure 5). Actu-ally, we also note that there are lots of atomicclouds below the Galactic plane in the similarregion (e.g., see the l = 22 ◦ − ◦ and b < ∼ − ◦ region in Figure 11). This feature probably in-dicates that the gas of the thick disk is proba-bly decoupled from the thin gas layer in someregions.Similarly, we also measured the mean thick-ness of the molecular disk in the whole rangeof l = 16 ◦ − ◦ (the upper panels in Figure 6)and l = 22 ◦ − ◦ (the lower panels in Figure 6)by using the best fit of the two Gaussian com-ponents. The thickness of the thin and thickCO disk is roughly 107 pc and 301 pc, respec-tively. Both of the two values from the widerlongitude region are ∼
15% larger than the re-sults from the smaller sub-regions of ∆ l = 6 ◦ (i.e., 107 pc vs. 94 pc for the thin CO disk and301 pc vs. 256 pc for the thick CO disk).We further investigate the variation of thethin disk over the Galactic longitude. Basedon the measurements of the 30 sub-regions with∆ l = 1 ◦ , the mean thickness of the thin COlayer (excluding l = 16 − ◦ ) is ∼
85 pc (see the right panel in Figure 7), which is slightlysmaller than values of ∼ l ∼ − ◦ , while the thicknessof the l ∼ − ◦ region is considerably smaller(see the right panel in Figure 7). According toFigure 3, we also note that the CO emission in l = 16 ◦ and l = 22 ◦ is relatively weak with re-spect to other regions near the tangent points.We speculated that it is probably due to the ef-fect of the dynamical perturbation within the3-kpc ring.The larger values of the thickness of the COlayer from Figure 6 can be attributed to thevariation of z at different Galactic longitude,which will widen the thickness of the gas diskby fitting data from the larger longitude range.Indeed, we find that z changes from negativevalues at l ∼ − ◦ to positive values at l ∼ − ◦ (see the left panel in Figure 7).As discussed in Heyer & Dame (2015), othermethods probably yield a higher value for theCO layer thickness due to corrugations in theinner Galactic disk (e.g., FWHM ∼ V ( l ) > ∼ V tan ( l ) − − and V ( l ) > ∼ V tan ( l ) −
10 km s − . The resultant fittingsare not changed significantly. To recapitulatebriefly, our recommended thickness of the thinCO layer of ∼
85 pc is in agreement with thevalue of ∼
90 pc from Malhotra (1994b). Andthe thickness of the thick CO layer is about260–300 pc, which is also consistent with pre-vious studies (e.g., Dame & Thaddeus 1994).The thick molecular disk revealed by the faintCO emission is about 3 times as wide as thewell-known central thin CO layer. The thick-ness of the thick CO layer is also comparable inwidth to the central H i layer (e.g., 250-270 pc inDickey & Lockman 1990; Lockman & Gehman1991).Moreover, the thickness of the thick CO layerfrom the MWISP survey is just between ∼
190 pc for the diffuse H gas and ∼
320 pc forthe diffuse H i +warm ionized medium (WIM)gas by considering R =8.15 kpc (see the discus-sions of the bright/faint diffuse C ii -without-COemission in Velusamy & Langer 2014). Theseresults show that a considerable amount ofmolecular gas in the CO-faint H clouds canbe revealed by the high-resolution and high-sensitivity CO survey. In the following section,we will investigate the distributions and prop-erties of the CO-faint molecular gas based onthe new identified high- z MCs near the tangentpoints.3.3.
MCs far from the Galactic Plane
Identification of the high- z MCs
MCs far from the Galactic plane are identi-fied by using the density-based spatial cluster-ing of applications with noise (DBSCAN ) clus-tering algorithm. Full details can be found inYan et al. (2020), and a brief description of themethod is presented below.First, the 3D PPV data cubes in the veloc-ity interval of 40–160 km s − are smoothed to0.5 km s − to decrease the random noise fluctu-ations in the velocity axis. In the PPV space,the minimum number of neighborhood voxels(MinPts) is set as 16 and the minimum cutoff onthe data is 2 × rms (i.e., the lowest level to sur-round the 3D structure). And the peak bright-ness temperature ( T peak ) is ≥ × rms (here therms is ∼ − for CO) to control the qualityof the selected samples. Second, the projectionarea on the spatial scale has at least 4 pixels ( ∼ one beam) and the minimum channel numberin the velocity axis is considered to be ≥ https://scikit-learn.org/stable/auto examples/cluster/plot dbscan.html ducing striping artefacts and other uncertaintiesin the whole data. These criteria are helpful toobtain as much of the emission as possible andto avoid the contamination of the noise in the3D data, leading to pick up weak but true sig-nal for the large-scale CO data. Broad criteria(e.g., MinPts=12 and/or T peak = 3 × rms) willincrease the number of the MC sample, how-ever, there is usually large uncertainty for thetrue MC identification due to the confusion withthe noise fluctuations.Finally, MCs far from the Galactic planeare selected based on the further criteria of V MC > ∼ V tan ( l ) − − and z MC ( l, b, v ) ≤ z ( l ) − σ z ( l ) or z MC ( l, b, v ) ≥ z ( l ) + 3 σ z ( l ).Here, z ( l ) and σ z ( l ) can be obtained based onthe best fit of z ( l )– l and σ z ( l )– l relations fromthe thin CO disk (see the measurements and thefitting lines in Figure 7). That is, MCs in thethin CO disk are excluded due to their concen-trated distribution in the region of b ∼ ◦ , i.e., z ( l ) − σ z ( l ) < z MC (thin disk) < z ( l ) + 3 σ z ( l ).In total, 1055 samples near the tangent pointswere identified as the MCs far from the Galac-tic plane between l = 16 ◦ and l = 52 ◦ based onthe MWISP CO data. The MCs, which haveweak CO emission, are relatively isolated withrespect to the molecular gas near the Galac-tic midplane. Table 1 lists the parameters ofeach MC, i.e, (1) the ID of the identified MCs,arranged from the low Galactic longitude; (2)and (3) the MC’s Galactic coordinates ( l and b ); (4) the MC’s LSR velocity ( V LSR ); (5) theMC’s one-dimensional velocity dispersion ( σ v );(6) the MC’s peak emission ( T peak ); (7) theMC’s area; (8) the MC’s distance obtained fromthe tangent points, i.e, d = R × cos( l ); (9)the MC’s z scale defined as z = d × tan( b );(10) the MC’s mass estimated from the CO-to-H conversion factor method, X CO = 2 × cm − (K km s − ) − (e.g., Dame et al. 2001;Bolatto et al. 2013); and (11) the MC’s virialparameter α = M virial M Xfactor = σ v RGM
Xfactor , here, R ertical Distribution of MCs G the gravita-tional constant. We can easily use MWISPGlll.lll ± bb.bbb ± vvv.vv to name an MC iden-tified from the CO survey.3.3.2. Properties and Statistics of the high- z MCs
For the molecular gas far from the Galacticplane, Figure 8 displays the vertical distribu-tion of the 1055 MCs, which can be fitted bya Gaussian function with z = − . σ z = 113 . ∼
113 pc (and thus theFWHM of ∼
270 pc) for the gas layer tracedby the ensemble of the high- z MCs, which isroughly similar to the measurement from theCO emission in the thick molecular gas layer(Section 3.1).Importantly, the excellent agreement of thevertical distribution of these discrete MCs withthat of the thick CO disk emission proposed byDame & Thaddeus (1994) confirms their claimthat the high- z emission they observed is notsubstantially contaminated by sidelobe pickupfrom the central disk. Therefore, our resultsbased on the large-scale, high-resolution, andhigh-sensitivity data demonstrate convincinglythat the thick molecular gas disk is indeed animportant component in the inner Galaxy. Thethick molecular disk is composed of many dis-crete MCs with small size and low mass (seebelow).Figure 9 shows the properties of the MCs.Typically, these MCs have T peak ∼ ∼ M ⊙ pc − assuming aconstant CO-to-H conversion factor) for thesesmall MCs. The effective radius of these MCsare 0 . ′ . ′ ∼ γ =-1.74 (i.e., N ( M > M ) ∝ M γ +1 forthe cloud’s mass range of ∼ − M ⊙ ) can describe the mass distribution of the high- z MCs, which is consistent to γ ∼ -1.7 ob-served for the MCs in the Milky Way (e.g.,Roman-Duval et al. 2010; Heyer & Dame 2015;Rice et al. 2016; Colombo et al. 2019). Thevelocity dispersion ( σ v ) of the high- z MCs isroughly 0.8 km s − , leading to α > ∼
10 for themost of CO clouds.Obviously, the lower limit of T peak , radius,and surface density is related to (1) the lim-ited sensitivity and spatial resolution of the cur-rent CO survey, and (2) the selection criterionof the MCs used in Section 3.3.1. We men-tion that some clouds with weak CO emission( T peak < ∼ < ∼ ′ )are missed due to the observational limit ofthe MWISP survey. Therefore, a substan-tial amount of the fainter CO emission (e.g., < ∼ − or M < ∼ − M ⊙ ) couldprobably be unveiled by higher-sensitivity COsurveys.The cloud-cloud velocity dispersion ( σ cc ) ofthese high- z MCs can be estimated from thesamples with V LSR > ∼ V tan . Obviously, the es-timation of σ cc depends on the definition of l − V tan relation. By taking into account theslight variation of the tangent velocity, we findthat the cloud-cloud dispersion varies from 4 . ± . − for V LSR > ∼ V tan to 5 . ± . − for V LSR > ∼ V tan − − with the 1 ◦ bin samplesin the l = [+16 ◦ , +52 ◦ ] region. The cloud-cloud velocity dispersion only changes a littlewith the variation of the selected tangent ve-locity. On the other hand, the estimated σ cc of these MCs is comparable to the measuredcloud-cloud velocity dispersion of the total MCsamples near the Galactic plane (see Section3.4 and other works, e.g., Stark & Brand 1989;Malhotra 1994b).Interestingly, the low σ cc of the high- z MCs in the Milky Way is different from thefinding that a potential thick gas disk mayhave the high velocity dispersion for galaxies(e.g., σ cc ∼
12 km s − for the thick molecu-lar gas disk with extended/diffuse CO emis-sion, Cald´u-Primo et al. 2013). As discussed inKrumholz et al. (2018), the high mass transportrates and star formation rates can lead to thehigh velocity dispersions of the gas in galaxies.Briefly, we summarized all of the physical prop-erties for the high- z MCs in Table 2.As seen in Figure 10, the MCs are mainlyconcentrated in regions of l = 22 ◦ − ◦ (i.e., R GC =3.1–4.1 kpc) and l = 43 ◦ − ◦ (i.e., R GC =5.6–6.4 kpc). And a smaller concentra-tion is between the two peak distributions (i.e.,see the l = 32 ◦ − ◦ or R GC =4.3–5.0 kpc re-gion in Figure 10). We suggest that the large-scale structures (i.e., the Norma arm, the AquilaSpur toward the tip of the Galactic bar, andthe Carina-Sagittarius arm, see Figures 1 and3, refer to Bland-Hawthorn & Gerhard 2016;Reid et al. 2016) near the tangent points areprobably responsible for the three peaks at thelongitude distribution (or R GC ) of the enhancedhigh- z MCs. Interestingly, the scatter of σ cc isrelatively larger in the region of l ∼ ◦ − ◦ (Figure 12), which indicates that the Galac-tic bar may have important impact on the dy-namics and distribution of the gas in the innerGalaxy.We also find that the high- z H i gas is abun-dant in the region of l ∼ ◦ − ◦ (Ford et al.2010), although the vertical distribution of thehigh- z H i clouds is much wider than that of theCO clouds (i.e., 300 pc < ∼ | z HI cloud | < ∼ < ∼ | z CO cloud | < ∼
250 pc in the 90% range).Figure 11 shows the comparison between thehigh- z MWISP CO clouds and the GASS H i clouds in the range of l=[+17 . ◦ . ◦
2] andb=[+5 . ◦ − . ◦ i sam-ples in Ford et al. (2010) are very likely relatedto our CO clouds, e.g., (1) GASS H i cloud:(18 . ◦ − . ◦
00, 126.1 km s − ) vs. MWISP COcloud: (18 . ◦ − . ◦ − ); (2)GASS H i cloud: (19 . ◦
19, +2 . ◦
06, 136.3 km s − ) vs. MWISP CO cloud: (19 . ◦ . ◦ − ); (3) GASS H i cloud: (19 . ◦ . ◦
43, 138.0 km s − ) vs. MWISP CO cloud:(19 . ◦ . ◦ − ), etc. The spa-tial and kinematical relationships between theH i cloud and the CO cloud are worth exploringin more detail in future works by using the high-quality H i data (e.g., the GALFA-HI survey,Peek et al. 2011, 2018) and MWISP CO data.Here we briefly discussed the possible origin ofthe high- z MCs in Section 3.3.3.In the region of | z − z | > σ z (i.e., the regionoutside the 1 × σ z of the thick CO disk), themolecular gas mass is about 3 . × M ⊙ basedon the discovered high- z MCs by adopting z =-15 pc and σ z =120 pc (e.g., see Figure 8). Thevalue is approximately 31 .
7% of the total massof the thick CO disk in the 5.1 kpc long redbelt (see Figure 1; including an extrapolationof the component through the full Galactic diskby assuming the Gaussian distribution). Thus,the total mass of the thick CO disk in the beltregion is about 1 . × M ⊙ . Adopting a meanwidth of 0.5 kpc for the red belt (i.e., the meanpath length along the line of sight for the belt,see the model in Lockman 1984), we obtain anarea of 2.55 kpc (i.e., 5.1 kpc length × ∼ . M ⊙ pc − , leading to the volume densityof ∼ . M ⊙ pc − (or ∼ .
02 H cm − ) in thediscussed region by assuming the Gaussian dis-tribution of ρ ( z ) = ρ exp( − ( z − z ) σ ). We men-tioned that the estimated values are probablythe lower limit due to the limited observationalsensitivity and resolution of the MWISP sur-vey (i.e., the missing molecular gas mass for thevery faint CO emission and small clouds thatare not involved in the calculation). In the re-gion of R GC =2–8.15 kpc, the molecular massof the thick CO disk is thus ∼ . × M ⊙ ,which is about 10% of the total molecular massin the inner Galaxy (i.e., ∼ . × M ⊙ for the ertical Distribution of MCs Origin of the high- z MCs
According to the vertical distribution of thehigh- z molecular gas (Figure 8), we find that theidentified MCs are mainly concentrated in the100 pc < ∼ | z | < ∼
360 pc region. However, thereis only little amount of molecular gas tracedby CO emission in the region of | z | > ∼
360 pc(i.e., 32/1055 high- z MCs). We thus sug-gest that these MCs with faint CO emissionprobably belong to the disk population. Thespatial and velocity coincidence between someCO clouds and the corresponding H i cloudsis very interesting. It indicates that consider-able amounts of molecular gas appear to sur-vive in the disk-halo transition region (or in thedisk-halo zone close to the Galactic midplane,e.g., 100 pc < ∼ | z | < ∼
300 pc). The coexistenceof molecular gas and the atomic gas in the disk-halo transition region need to be further investi-gated based on the observational and theoreticalstudies.Generally, the typical internal crossing timeof these high- z MCs is about 3 Myr assumingthe typical radius of ∼ σ v = 0 . − (Figure 9). This value isroughly an order of magnitude smaller than thedynamical timescale of the moving clouds fromthe Galactic midplane (i.e., t dyn = zσ cc > ∼
24 Myr,here σ cc = 4 . − is the cloud-cloud veloc-ity dispersion from Section 3.4). It indicatesthat the MCs’ memory at their birthplace islost. And any observed kinematical featuresfrom the line profile and spatial morphologyprobably represent the turbulence in the cur-rent local environment. Considering the preva-lent turbulence in the ISM, the cloud will bedestroyed quickly due to the Kelvin-Helmholtzand Rayleigh-Taylor instabilities (e.g, < ∼ z MCs are un- stable unless they are confined by some externalpressure. The low σ cc of ∼ − and large σ z of 260–300 pc for these high- z MCs indicatesthat the gravitational force of the midplane isnot balanced by the gas pressure, suggestingthat the ensemble of the MCs is in general outof equilibrium. All of these results show thatthe high- z MCs probably have short lifetimeless than several Myr (e.g., comparable to thetypical internal crossing time). Indeed, if theMCs directly move from the Galactic plane totheir current places, the moving velocity of theclouds should be an order of magnitude largerthan that of the cloud-cloud dispersion veloc-ity because of the short lifetime of these clouds(e.g., < ∼ several Myr). Alternatively, some ofthe high- z MCs are probably newborn cloudsdue to the rapid H formation in shock inter-action regions (e.g., with the high ram pres-sure of > ∼ − K cm − , see Su et al. 2018).We suggest that these MCs are transitory. Thehigh- z molecular gas with faint CO emission iseither dispersing or being assembled by someexternal dynamical processes (e.g., compressionby shocks, cloud-cloud collision, and shear mo-tions, etc).Whatever the exact formation mechanism ofthe high- z clouds, the energetic sources near theGalactic plane may play important roles in theorigin of the disk-population gas (e.g., the stel-lar feedback from massive stellar winds and/orsupernova explosions, see the Galactic fountainmodels in Shapiro & Field 1976; Bregman 1980;Houck & Bregman 1990; Spitoni et al. 2008;Soler et al. 2020). Recently, di Teodoro et al.(2020) also found that the cold, dense, and high-velocity molecular gas survives in the MilkyWay’s nuclear wind at ∼ ∼ i (C ii ) gas(e.g., ∼ i emission and ∼ ii emission, Dickey & Lockman1990; Lockman & Gehman 1991; Langer et al.2014; Velusamy & Langer 2014) may be a rel-evant hint to establish a link between the for-mation/evolution of the high- z MCs and var-ious physical processes. Further studies willbe helpful to clarify these issues by a combi-nation of multi-wavelength data, e.g., kinemat-ical connection between the CO clouds and theH i clouds, possible related IR features, and/oremission from other tracers like OH 18 cm, CH3.3 GHz lines, and 158 µ m [C ii ] lines, etc.3.4. Cloud-to-Cloud Velocity Dispersion nearthe Tangent Points and the TotalMid-plane Density
The H i and H gas in the inner Galaxy( R GC = R < ∼ .
15 kpc) is concentrated towardthe thin plane due to the gravitational poten-tial of the matter in the Galactic disk. If thepressure from magnetic fields, cosmic rays, andthe radiation field is ignored, the vertical dis-tribution of the H gas is mainly determinedby the total gravitating mass near the disk (i.e.,stars, gas, and possible other unknown objects).Considering the balance between the turbulentpressure of the isothermal molecular gas andthe gravitational force on the Galactic plane,we can obtain the formula of ρ = σ πGσ forthe Gaussian distribution of the gas layer (e.g.,see Malhotra 1994b, 1995). Here ρ is the totaldensity of the midplane mass, σ cc is the cloud-cloud velocity dispersion, and σ z is the scaleheight of the thin CO layer. And we ingorethe scale height of the thick CO disk due to itslimited contribution to the total molecular gas mass (i.e., ∼
3% for the | z | > ∼
120 pc moleculargas).In principle, we can estimate the midplanemass density of the Galactic disk if σ cc and σ z are obtained from observations. Figure 12displays the distribution of σ cc for all MCsnear the tangent points (i.e., MCs in the thinand thick disk with V LSR (MC) > ∼ V tan − ∆V).Here ∆ V is simply selected as 3 . − (i.e.,roughly three times of the scatter of σ cc , whichis nearly constant for the tangent MCs in thelongitude range) by taking into account the tan-gent MCs with somewhat lower V LSR . We notethat the estimated value of σ cc is in the rangeof 2.7–7.9 km s − (i.e., mean value of 4.9 km s − and a scatter of 1.3 km s − ), which is close tothat of the high- z MCs discussed in Section3.3.2. The value is roughly consistent withother works based on different approaches (e.g.,see Blitz et al. 1984; Clemens 1985; Stark 1984;Stark & Brand 1989; Malhotra 1994b).Additionally, the distribution of σ cc seemsto display a linear variation with a slope of ∼ − . − kpc − between R GC =2.2–6.4 kpc.It is interesting to note that the slope from theCO data is about one half of the H i gas whenthe joint gravitational potential is considered(i.e., stars, H i gas, and H gas near the disk;see Narayan & Jog 2002). The decrease of σ cc at larger R GC may be explained by a lower starformation rate in the outer part of the innerdisk if σ cc (i.e., the turbulence of the MCs) isregulated by the energy input via star-formingactivities near the Galactic plane. On the otherhand, however, it should be mentioned that themeasured σ cc based on our CO data also dis-plays considerable fluctuation in different R GC (see Figure 12), which is probably related to thedifferent localized environment. More quantita-tive studies are required to better understandthese features.Combining the measurements of R GC – σ cc and R GC – σ z for MCs near the tangent points, we ertical Distribution of MCs R GC – ρ inFigure 13). The error of ρ can be calculatedfrom the fitting error of σ z (Figure 7) and the as-sumed error of σ cc (Figure 12). We find that anexponential-disk model of ρ ( R ) = ρ GC e − R/R sl can roughly fit the distribution of the totalmass density for the R GC =2.25–6.42 kpc re-gion. Here, ρ GC =1.28 M ⊙ pc − is the mass den-sity at the Galactic center and R sl = 3 .
20 kpcis the scalelength of the disk. The midplanemass density at the Sun is estimated to be ρ ( R )=0.10 M ⊙ pc − , which agrees well withthe Oort limit summarized in McKee et al.(2015). Because of the lower CO emission inthe tangent points of l < ∼ ◦ (i.e., R GC < ∼ R GC =3.05–6.42 kpc region. Theresultant fitting of R GC – ρ gives the similar re-sult (see the blue line in Figure 13). SUMMARYBased on the MWISP CO data, we have per-formed a study of the vertical distribution ofthe molecular gas near the tangent points in therange of l = 16 ◦ –52 ◦ and | b | < ∼ . ◦
1. The main re-sults are:1. The terminal velocity of the molecular gas,which is comparable to other observations andtheoretical models, is well determined based onthe new MWISP CO survey toward the innerGalaxy (Figure 2).2. We use the CO emission near the terminalvelocity to trace the molecular gas distributionat the tangent points, which can avoid the dis-tance ambiguity within the solar circle. Thehigh-quality CO data reveal two main molec-ular gas features near the tangent points, in-cluding large-scale bright CO emission in theGalactic midplane and discrete MCs with veryweak emission in relatively high- b regions (seeFigure 3).3. Based on the CO integrated intensity nearthe terminal velocity, we find that the model of two Gaussian components can be used to de-scribe the vertical distribution of the moleculargas near the tangent points. The narrow onewith a FHWM of ∼
85 pc is the well-known thinmolecular gas disk in the inner Galaxy, while an-other broad component with a FHWM of ∼ z ) and the scaleheight ( σ z ) of the molecular gas with respect tothe Galactic longitude, which is consistent withprevious studies (e.g., Malhotra 1994b). For thethick CO disk, its thickness is about 3 timeswide of that of the well-known thin moleculargas layer. And the thickness of the thick molec-ular gas layer is well comparable in width tothe central atomic gas layer (e.g., 250-270 pc inDickey & Lockman 1990; Lockman & Gehman1991).4. For the thick CO disk, a total of 1055 high- z MCs were identified at the tangent points inthe first quadrant region of the MWISP sur-vey. These MCs have a median radius of 2.5 pc,a median velocity dispersion of 0.8 km s − ,and a median peak temperature of 2.1 K. Themass surface brightness of the MCs are verylow, i.e., a median value of 6.8 M ⊙ pc − as-suming a constant CO-to-H conversion factorof 2 × cm − (K km s − ) − . The cloud-cloud velocity dispersion of the high- z MCs is4.4–5.6 km s − , which is similar to the valueof 4.9 ± . − for all MCs at the tangentpoints. The high virial parameter indicates thatthe high- z MCs are probably short-lived ob-jects. That is, the MCs of the new disk pop-ulation are either dispersing or being assembledby some external dynamical processes. Alter-natively, some of the high- z MCs are probablynewborn clouds in the environment of the highram pressure. Our findings show that the thickCO disk is composed of many discrete MCs withsmall size and low mass (i.e., dNdM ∝ M − . inthe MC’s mass range of ∼ − M ⊙ ).25. Nearly 90% of these MCs are > ∼
100 pc fromthe plane. However, only 3% of MCs are in the | z | > ∼
360 pc region, suggesting that the high- z molecular gas is the disk population. Includ-ing the emission of the CO layer through thewhole Galactic plane, the total molecular gasmass of the thick CO disk is estimated to be8 . × M ⊙ in the range of R GC =2–8.15 kpc,which is about 10% of the total molecular massin the same region of the inner Galaxy. The sur-face density and midplane mass density of thethick molecular gas disk are ∼ . M ⊙ pc − and ∼ . M ⊙ pc − (or ∼ .
02 H cm − ), respec-tively.6. The cloud-cloud velocity dispersion of alltangent MCs seems to be smaller at larger R GC ,leading to a slope of ∼ − . − kpc − in theregion of R GC =2.2–6.4 kpc. This value is aboutone half of the H i gas when the joint gravita-tional potential is considered (Narayan & Jog2002). The higher star formation rate at smaller R GC is probably responsible for this feature.7. Assuming the vertical equilibrium be-tween the turbulent pressure of the molec-ular gas and the total gravitational force in the disk, we can use an exponential-diskmodel of ρ ( R ) = ρ GC e − R/R sl to fit the dis-tribution of the mass density in the Galac-tic midplane. The best-fit parameters are R sl =3.2 kpc and ρ GC =1.28 M ⊙ pc − . Themidplane mass density in the solar neigh-borhood is estimated to be 0.10 M ⊙ pc − at R =8.15 kpc, which agrees with the local massdensity of 0.097 ± . M ⊙ pc − suggested byMcKee et al. (2015).ACKNOWLEDGMENTSWe are very grateful for the support pro-vided by the staff members of Qinghai Ra-dio Observing Station at Delingha. We alsothank the anonymous referee for providingmany helpful comments and suggestions thatlargely improved the paper. MWISP is fundedby the National Key R&D Program of China(2017YFA0402700) and the Key Research Pro-gram of Frontier Sciences of CAS (QYZDJ-SSW-SLH047). Y.S. is supported by the NSFCgrant No. 11773077. X.C. acknowledges sup-port by the NSFC through grant No. 11629302. Facility:
PMO 13.7m
Software:
GILDAS/CLASS (Pety 2005)REFERENCES
Benedettini, M., Molinari, S., Baldeschi, A., et al.2020, A&A, 633, A147,doi: 10.1051/0004-6361/201936096Bland-Hawthorn, J., & Gerhard, O. 2016,ARA&A, 54, 529,doi: 10.1146/annurev-astro-081915-023441Blitz, L., Magnani, L., & Mundy, L. 1984, ApJL,282, L9, doi: 10.1086/184293Bolatto, A. D., Wolfire, M., & Leroy, A. K. 2013,ARA&A, 51, 207,doi: 10.1146/annurev-astro-082812-140944Bregman, J. N. 1980, ApJ, 236, 577,doi: 10.1086/157776Bronfman, L., Cohen, R. S., Alvarez, H., May, J.,& Thaddeus, P. 1988, ApJ, 324, 248,doi: 10.1086/165892 Cald´u-Primo, A., Schruba, A., Walter, F., et al.2013, AJ, 146, 150,doi: 10.1088/0004-6256/146/6/150Clemens, D. P. 1985, ApJ, 295, 422,doi: 10.1086/163386Clemens, D. P., Sanders, D. B., & Scoville, N. Z.1988, ApJ, 327, 139, doi: 10.1086/166177Colombo, D., Rosolowsky, E., Duarte-Cabral, A.,et al. 2019, MNRAS, 483, 4291,doi: 10.1093/mnras/sty3283Combes, F. 1991, ARA&A, 29, 195,doi: 10.1146/annurev.aa.29.090191.001211Dame, T. M., Hartmann, D., & Thaddeus, P.2001, ApJ, 547, 792, doi: 10.1086/318388Dame, T. M., & Thaddeus, P. 1994, ApJL, 436,L173, doi: 10.1086/187660 ertical Distribution of MCs —. 2008, ApJL, 683, L143, doi: 10.1086/591669—. 2011, ApJL, 734, L24,doi: 10.1088/2041-8205/734/1/L24di Teodoro, E. M., McClure-Griffiths, N. M.,Lockman, F. J., & Armillotta, L. 2020, Nature,584, 364. https://arxiv.org/abs/2008.09121Dickey, J. M., & Lockman, F. J. 1990, ARA&A,28, 215,doi: 10.1146/annurev.aa.28.090190.001243Du, X., Xu, Y., Yang, J., & Sun, Y. 2017, ApJS,229, 24, doi: 10.3847/1538-4365/aa5d9dDu, X., Xu, Y., Yang, J., et al. 2016, ApJS, 224,7, doi: 10.3847/0067-0049/224/1/7Eden, D. J., Moore, T. J. T., Currie, M. J., et al.2020, MNRAS, 498, 5936,doi: 10.1093/mnras/staa2734Ford, H. A., Lockman, F. J., & McClure-Griffiths,N. M. 2010, ApJ, 722, 367,doi: 10.1088/0004-637X/722/1/367Ford, H. A., McClure-Griffiths, N. M., Lockman,F. J., et al. 2008, ApJ, 688, 290,doi: 10.1086/592188Garcia-Burillo, S., Guelin, M., Cernicharo, J., &Dahlem, M. 1992, A&A, 266, 21Heyer, M., & Dame, T. M. 2015, ARA&A, 53,583, doi: 10.1146/annurev-astro-082214-122324Houck, J. C., & Bregman, J. N. 1990, ApJ, 352,506, doi: 10.1086/168554Iwasaki, K., Tomida, K., Inoue, T., & Inutsuka,S.-i. 2019, ApJ, 873, 6,doi: 10.3847/1538-4357/ab02ffKrumholz, M. R., Burkhart, B., Forbes, J. C., &Crocker, R. M. 2018, MNRAS, 477, 2716,doi: 10.1093/mnras/sty852Langer, W. D., Pineda, J. L., & Velusamy, T.2014, A&A, 564, A101,doi: 10.1051/0004-6361/201323281Lockman, F. J. 1984, ApJ, 283, 90,doi: 10.1086/162277—. 2002, ApJL, 580, L47, doi: 10.1086/345495Lockman, F. J., & Gehman, C. S. 1991, ApJ, 382,182, doi: 10.1086/170706Malhotra, S. 1994a, ApJ, 437, 194,doi: 10.1086/174988—. 1994b, ApJ, 433, 687, doi: 10.1086/174677—. 1995, ApJ, 448, 138, doi: 10.1086/175946McClure-Griffiths, N. M., & Dickey, J. M. 2016,ApJ, 831, 124,doi: 10.3847/0004-637X/831/2/124 McKee, C. F., Parravano, A., & Hollenbach, D. J.2015, ApJ, 814, 13,doi: 10.1088/0004-637X/814/1/13Nakanishi, H., & Sofue, Y. 2006, PASJ, 58, 847,doi: 10.1093/pasj/58.5.847Narayan, C. A., & Jog, C. J. 2002, A&A, 394, 89,doi: 10.1051/0004-6361:20021128Peek, J. E. G., Heiles, C., Douglas, K. A., et al.2011, ApJS, 194, 20,doi: 10.1088/0067-0049/194/2/20Peek, J. E. G., Babler, B. L., Zheng, Y., et al.2018, ApJS, 234, 2,doi: 10.3847/1538-4365/aa91d3Pety, J. 2005, in SF2A-2005: Semaine del’Astrophysique Francaise, ed. F. Casoli,T. Contini, J. M. Hameury, & L. Pagani, (LesUlis: EDP), 721Pety, J., Schinnerer, E., Leroy, A. K., et al. 2013,ApJ, 779, 43, doi: 10.1088/0004-637X/779/1/43Putman, M. E., Peek, J. E. G., & Joung, M. R.2012, ARA&A, 50, 491,doi: 10.1146/annurev-astro-081811-125612Reid, M. J., Dame, T. M., Menten, K. M., &Brunthaler, A. 2016, ApJ, 823, 77,doi: 10.3847/0004-637X/823/2/77Reid, M. J., Menten, K. M., Brunthaler, A., et al.2019, ApJ, 885, 131,doi: 10.3847/1538-4357/ab4a11Rice, T. S., Goodman, A. A., Bergin, E. A.,Beaumont, C., & Dame, T. M. 2016, ApJ, 822,52, doi: 10.3847/0004-637X/822/1/52Rigby, A. J., Moore, T. J. T., Plume, R., et al.2016, MNRAS, 456, 2885,doi: 10.1093/mnras/stv2808Roman-Duval, J., Jackson, J. M., Heyer, M.,Rathborne, J., & Simon, R. 2010, ApJ, 723,492, doi: 10.1088/0004-637X/723/1/492Sanders, D. B., Solomon, P. M., & Scoville, N. Z.1984, ApJ, 276, 182, doi: 10.1086/161602Schuller, F., Csengeri, T., Urquhart, J. S., et al.2017, A&A, 601, A124,doi: 10.1051/0004-6361/201628933Schuller, F., Urquhart, J. S., Csengeri, T., et al.2021, MNRAS, 500, 3064,doi: 10.1093/mnras/staa2369Shan, W., Yang, J., Shi, S., et al. 2012, ITTST, 2,593, doi: 10.1109/TTHZ.2012.2213818Shapiro, P. R., & Field, G. B. 1976, ApJ, 205,762, doi: 10.1086/154332 Sofue, Y., Kohno, M., Torii, K., et al. 2019, PASJ,71, S1, doi: 10.1093/pasj/psy094Soler, J. D., Beuther, H., Syed, J., et al. 2020,A&A, 642, A163,doi: 10.1051/0004-6361/202038882Spitoni, E., Recchi, S., & Matteucci, F. 2008,A&A, 484, 743,doi: 10.1051/0004-6361:200809403Stark, A. A. 1984, ApJ, 281, 624,doi: 10.1086/162137Stark, A. A., & Brand, J. 1989, ApJ, 339, 763,doi: 10.1086/167334Su, Y., Sun, Y., Li, C., et al. 2016, ApJ, 828, 59,doi: 10.3847/0004-637X/828/1/59Su, Y., Zhou, X., Yang, J., et al. 2018, ApJ, 863,103, doi: 10.3847/1538-4357/aad04eSu, Y., Yang, J., Zhang, S., et al. 2019, ApJS, 240,9, doi: 10.3847/1538-4365/aaf1c8 Sun, J. X., Lu, D. R., Yang, J., et al. 2018,AcASn, 59, 3Sun, Y., Su, Y., Zhang, S.-B., et al. 2017, ApJS,230, 17, doi: 10.3847/1538-4365/aa63eaSun, Y., Xu, Y., Yang, J., et al. 2015, ApJL, 798,L27, doi: 10.1088/2041-8205/798/2/L27Sun, Y., Yang, J., Xu, Y., et al. 2020, ApJS, 246,7, doi: 10.3847/1538-4365/ab5b97Torii, K., Fujita, S., Nishimura, A., et al. 2019,PASJ, 71, S2, doi: 10.1093/pasj/psz033Vall´ee, J. P. 2014, ApJS, 215, 1,doi: 10.1088/0067-0049/215/1/1Velusamy, T., & Langer, W. D. 2014, A&A, 572,A45, doi: 10.1051/0004-6361/201424350Yan, Q.-Z., Yang, J., Su, Y., Sun, Y., & Wang, C.2020, ApJ, 898, 80,doi: 10.3847/1538-4357/ab9f9c ertical Distribution of MCs Figure 1.
Face-on view of the Milky Way (R. Hurt: NASA/JPL-Caltech/SSC) superposed on the tangentpoint molecular gas of the Galaxy. The MCs near the tangent points are identified based on the MWISPCO data toward the longitude range of 16 ◦ –52 ◦ (or a red belt with a length of ∼ Figure 2.
Longitude–velocity diagram of the CO emission for the molecular gas along l = 16 ◦ − ◦ .The region has a width of 3 . ◦
0, i.e., b = [ − . ◦
5, 1 . ◦ V tan ) based on the MWISP CO data. The purple and golden lines show the rotationcurve from Reid et al. (2019) and the H i result (the thick line for the measured values and the thin line forthe linear fit) from McClure-Griffiths & Dickey (2016), respectively. ertical Distribution of MCs Figure 3. CO ( J =1–0) intensity map toward l = 16 ◦ − ◦ in the velocity range of V ( l ) > ∼ V tan ( l ) − − . The tangent velocity, i.e., V tan ( l ), is well determined from the red line in Figure 2. Some large-scalestructures seen in the longitude–velocity diagram (Figure 2) are also labeled on the map of the integratedCO emission near the tangent points (e.g., see Vall´ee 2014; Reid et al. 2016). -200 -100 0 100 200z (pc)0.00.20.40.60.81.01.2 I ( z ) /I pea k (a) -200 -100 0 100 200z (pc)0.00.20.40.60.81.01.2 I ( z ) /I pea k (b) -200 -100 0 100 200z (pc)0.00.20.40.60.81.01.2 I ( z ) /I pea k (c) -200 -100 0 100 200z (pc)0.00.20.40.60.81.01.2 I ( z ) /I pea k (d) -200 -100 0 100 200z (pc)0.00.20.40.60.81.01.2 I ( z ) /I pea k (e) -200 -100 0 100 200z (pc)0.00.20.40.60.81.01.2 I ( z ) /I pea k (f) Figure 4. (a)–(f) Vertical distribution of the total integrated CO (black solid line) and CO (black solidline, scale to 0.5) intensity toward the longitude range of 16 ◦ − ◦ , 22 ◦ − ◦ , 28 ◦ − ◦ , 34 ◦ − ◦ , 40 ◦ − ◦ ,and 46 ◦ − ◦ , respectively. The bin in each panel is 5 pc, which is corresponding to about 5 pixels (2 . ′ ∼ CO emission, while the red dashed line is for the CO gas. Thefitted parameters of z and σ z (i.e., FWHM=2.355 × σ z ) are also labeled on each panel for the one Gaussiancomponent. ertical Distribution of MCs -400 -200 0 200 400z (pc)05101520 M ean I n t en s i t y ( K k m / s ) pe r b i n (a) -400 -200 0 200 400z (pc)0102030405060 M ean I n t en s i t y ( K k m / s ) pe r b i n (b) -400 -200 0 200 400z (pc)01020304050 M ean I n t en s i t y ( K k m / s ) pe r b i n (c) -400 -200 0 200 400z (pc)010203040 M ean I n t en s i t y ( K k m / s ) pe r b i n (d) -400 -200 0 200 400z (pc)0102030 M ean I n t en s i t y ( K k m / s ) pe r b i n (e) -400 -200 0 200 400z (pc)0102030 M ean I n t en s i t y ( K k m / s ) pe r b i n (f) Figure 5. (a)–(f) Vertical distribution of the mean intensity (i.e., the black solid line for I total / p N pixel ,see the text) of the CO emission toward the longitude range of 16 ◦ − ◦ , 22 ◦ − ◦ , 28 ◦ − ◦ , 34 ◦ − ◦ ,40 ◦ − ◦ , and 46 ◦ − ◦ , respectively. The bin in each panel is 5 pc. We assume p N pixel errors for each bin.The solid red line displays the best fitting of two Gaussian components (i.e., the fixed narrow dashed-linefrom Figure 4 + the fitted broad dashed-line) for the CO gas. The fitted parameters of z and σ z (i.e.,FWHM=2.355 × σ z ) are also labeled on each panel for the broad Gaussian component. Note that the zero-point of the fitting is roughly 4 K km s − , which is ∼ I rms of the CO integrated intensity near the tangentpoints. -200 -100 0 100 200z (pc)0.00.20.40.60.81.01.2 I ( z ) /I pea k (a) -400 -200 0 200 400z (pc)0102030 M ean I n t en s i t y ( K k m / s ) pe r b i n (b) -200 -100 0 100 200z (pc)0.00.20.40.60.81.01.2 I ( z ) /I pea k (c) -400 -200 0 200 400z (pc)0102030 M ean I n t en s i t y ( K k m / s ) pe r b i n (d) Figure 6.
Left panels: Same as Figure 4 but for the longitude range of 16 ◦ − ◦ and 22 ◦ − ◦ based onthe total integrated CO emission. The fitted parameters of the CO emission for the thin disk are used tofix the narrow Gaussian component in the right panels. Right panels: Same as Figure 5 but for the longituderange of 16 ◦ − ◦ and 22 ◦ − ◦ based on the mean intensity of the CO emission. The fitted parametersof z and σ z (i.e., FWHM=2.355 × σ z ) are labeled on each panel for the broad Gaussian component (i.e., thethick CO disk). ertical Distribution of MCs
50 40 30 20Longitude (deg)-60-40-2002040 z ( p c ) (a)
50 40 30 20Longitude (deg)1020304050 σ z ( p c ) (b) Figure 7.
The position of the centroid of the molecular gas near the tangent points ( z in the left panel)and the scale height of the thin gas disk ( σ z in the right panel) based on the CO emission at differentGalactic longitudes of l = 16 ◦ − ◦ with a separation of 1 ◦ . The line indicates the least-square fit from athird order polynomial by using the 36 points. -400 -200 0 200 400050100150200 -400 -200 0 200 400Z (pc)050100150200 N u m be r Gaussian_fit: (z =-15.9 pc, σ z =113.1 pc) Figure 8.
Vertical distribution of the identified MCs far from the Galactic plane. The red dashed-lineshows the Gaussian fit for the broad component of the Galactic thick disk. ertical Distribution of MCs peak (K)0100200300 N u m be r Mean=2.5 KMedian=2.1 K (a) σ v (km s -1 )050100150200250300 N u m be r Mean=0.9 km s -1 Median=0.8 km s -1 (b) N u m be r Mean=3.0 pcMedian=2.5 pc (c) xfactor (M sun )050100150200250300 N u m be r Mean=442.6 M sun
Median=136.3 M sun (d) sun pc -2 )050100150200 N u m be r Mean=7.9 M sun pc -2 Median=6.8 M sun pc -2 (e) N u m be r Mean=15.6Median=12.9 (f)
Figure 9. (a)–(f) Histogram of the peak temperature, the velocity dispersion, the radius, the mass, thesurface density, and the virial parameter of the CO cloud far from the Galactic plane.
20 30 40 50020406080100120 20 30 40 50Longitude (Degree)020406080100120 N u m be r (a)
60 80 100 120 140050100150 60 80 100 120 140Velocity (km s -1 )050100150 N u m be r (b) N u m be r (c) GC (kpc)020406080100 N u m be r (d) Figure 10. (a)–(d) Longitude, velocity, distance, and Galactocentric distance distributions of the MCs farfrom the Galactic plane. ertical Distribution of MCs Figure 11.
Comparison between the high- z CO clouds (crosses with the uniform size, this paper) and thedisk-halo H i clouds (circles, Ford et al. 2010). Note that the circle’s size is not the true angular size of theH i clouds, but it is proportional to the mass of the H i clouds.
50 40 30 20Longitude (deg)246810 V e l o c i t y D i s pe r s i on ( k m s - ) Mean=4.9 km s -1 Standard Deviation=1.3 km s -1 Figure 12.
Cloud-cloud velocity dispersion ( σ cc ) for all MCs (i.e., the thin + thick CO disk) near thetangent points toward l = 16 ◦ − ◦ . The line indicates the least-square fit from a third order polynomial(nearly linear form with a slope of ∼ − . − kpc − ) by using the current data of R GC ∼ ◦ is assumed as σ cc / √ N cloud . The mean value of σ cc and its standard deviationare also labeled on the panel. ertical Distribution of MCs GC (kpc)0.010.101.0010.00 D en s i t y ( M s un p c - ) ρ (R )=0.10 M sun pc -3 from R GC =6.42-2.25 kpc data ρ (R )=0.10 M sun pc -3 from R GC =6.42-3.05 kpc data Figure 13.
Midplane mass density for different Galactocentric distances ( R GC ) derived from the CO scaleheight (Figure 7b) and the velocity dispersion (Figure 12). The line shows the exponential disk model of ρ ( R ) = ρ GC e − R/R sl , where the fitted mass density at the Galactic center, ρ GC = 1 . M ⊙ pc − , and thescalelength, R sl = 3 .
20 kpc. The midplane mass density at the Sun ( R GC = R =8.15 kpc) is estimated tobe ∼ . M ⊙ pc − . Table 1.
Parameters of 1055 Molecular Clouds Far From the Galactic Plane based on the MWISP CO( J =1–0)EmissionID(1) l ( ◦ )(2) b ( ◦ )(3) V LSR ( km s − )(4) σ v ( km s − )(5) T peak (K)(6) Area(arcmin )(7) Distance a (kpc)(8) z scale(pc)(9) Mass( M ⊙ )(10) α b (11)1 16.103 0.896 143.16 2.07 1.84 10.75 7.8 122.5 690 31.52 16.278 0.976 137.99 1.00 1.29 2.50 7.8 133.2 100 24.23 16.436 -0.770 128.18 0.85 1.97 5.00 7.8 -105.0 160 15.44 16.498 0.809 130.07 1.31 2.60 14.25 7.8 110.4 620 15.95 16.526 0.929 139.87 3.27 2.54 15.75 7.8 126.7 1160 56.46 16.768 -0.600 128.00 1.37 3.30 31.50 7.8 -81.8 1700 9.57 16.801 -1.173 128.58 0.77 1.70 6.50 7.8 -159.8 220 10.5 Note — a The error of the estimated distance is about 2%–20% from l = 16 ◦ − ◦ assuming that the MCs arelocated near tangent points with a velocity uncertainty of ∼ − along the LOS (e.g., refer to the A5 model inReid et al. 2019). b The MC’s virial parameter estimated from the definition of α = σ v RGM
Xfactor (see Section 3.3.1).
Table 2.
Statistical Properties of the high- z Molecular Clouds a Median Mean T peak (K) Radius(pc) σ v ( km s − ) Mass( M ⊙ ) Surface Density( M ⊙ pc − ) α b σ cc ( km s − ) Thickness(pc) γ c ∼ ∼
280 -1.74
Note — a The properties of the molecular gas indicate that the high- z MCs belong to a new disk populationthat was not discovered by previous CO observations due to the low sensitivity, low resoltuion, and/or limitedlatitude coverage. b α is the virial parameter of the MCs (see the text). c γ is the spectral index of the massdistribution of the high- z MCs with a form dNdM ∝ M γ in the MC’s mass range of ∼ − M ⊙ . ertical Distribution of MCs Table 3.
Thickness of the Inner Galaxy Traced by Different TracersTracer FWHM a References b CommentsCO ∼
85 pc 1,2 The well-know thin molecular gas disk ∼
280 pc 1 The thick disk is composed of many discrete MCs with small size and low massC ii ∼
120 pc 3 C ii with bright CO emission as the dense H gas (the overestimated thin CO disk) ∼
190 pc 3 Bright diffuse C ii emission as the CO-faint diffuse H gas ∼
320 pc 3 Faint diffuse C ii emission as the diffuse H i and WIM gasH i ∼
230 pc 4,5 The narrow Gaussian component for the Cold Neutral Medium ∼
540 pc 4,5 The broad Gaussian component for the Warm Neutral Medium ∼ Note — a The estimated thickness has been scaled to R =8.15 kpc (e.g., Reid et al. 2019). bb