Gaia DR1 Evidence of Disrupting Perseus

Home , Perseus Arm

arXiv:1712.04714v2 [astro-ph.GA] 16 Jan 2018 ihlke l 05 idge ta.21) ute al. et Hunt 2016). al. et Lindegren (TGAS; 2015; solution al. astrometric et catalogue, Tycho-Gaia Tycho-2 Michalik the the with as mil- common two known in for mission stars parallaxes bright and 2016a), Gaia lion al. motions The et proper its spi- the Collaboration published including Gaia a recently (DR1; 2015). release 2016b) data of al. first al. et et side Hunt its Collaboration regarding also (Gaia leading information (see crucial and (2014) proper- provide origin al. trailing would al. velocity et arm et Kawata the stel- stellar (Grand ral between the and arms 2013). comparing ties gas spiral al. that et the characteristic Baba suggested by a 2016; affected 2012b, be motion should lar there that tr teeyradii, every at 2013; al. stars et 2015). Baba Baba 2012b,a; 2013; al. al. et et Grand D’Onghia 2011; al. tran- scale, et time Fujii rotating, short relatively differentially a ∼ on are patterns recurrent arm) arms sient, DYN (hereafter spiral (i.e., theory suggests spiral patterns dynamic wave hand, other long-lived rotating, & spi- rigidly characterises den- as arm) life- quasi-stationary SDW rals The different (hereafter dif- theory have 2014). wave two which sity Baba are & arms there (Dobbs spiral times galaxies of as- disk galactic theories in isolated ferent question For long-standing a tronomy. been has tained 5 0 y Slwo aleg18;Bb ta.2009; al. et Baba 1984; Carlberg & (Sellwood Myr 100 etu u srnmedrUiesttHiebr,Astron Heidelberg, Universitat der Astronomie fur Zentrum u.aanoa.p [email protected] [email protected]; eas Y r sams orttn ihthe with co-rotating almost is arm DYN a Because main- and created are galaxies disk in arms spiral How rpittpstuigL using 2018 typeset 17, Preprint January version Draft y;Ln&Su16;Bri i 96.O the On 1996). Lin & Bertin 1964; Shu & Lin Gyr; 1 uih Baba Junichi si h irpinpaeo rnin arm. transient a a favor of headings: strongly dyna phase Subject they disruption transient trends, the a qualitative in on to is based limited those are and comparisons scenario spiral density-wave ieteiec httevre eito rudtePresarm Perseus our the with around trends deviation observational vertex these neighborhood the solar compared the that in evidence deviation vertex direct positive the to contrast eas on eaievre eito o h ehiso h tr the on Cepheids inward the moving for and deviation d an faster vertex their radial rotating negative with is Galactocentric a correlated side are found the trailing arm also Both the We Perseus as the of arm, 1. Perseus kpc Release 1.5 Data within d Cepheids Gaia measured accurately from the of motions advantage taking variables, Cepheid 2 ulr pc cec aoaoy nvriyCleeLond College University Laboratory, Science Space Mullard ehv icvrdacersg ftedsuto hs ftePerse the of phase disruption the of sign clear a discovered have We 4 3 edlegrIsiu ¨rTertsh tde,Schloss Studien, für Theoretische Institut Heidelberger 6 eateto srnm,TeUiest fTko -- Hon 7-3-1 Tokyo, of University The Astronomy, of Department ulpIsiuefrAtooyadAtohsc,Universi Astrophysics, and Astronomy for Institute Dunlap 1. 1 Rcie eebr1,21;RvsdJnay1,21;Acce 2018; 12, January Revised 2017; 14, December (Received ask Kawata Daisuke , A INTRODUCTION T N E tl mltajv 01/23/15 v. emulateapj style X bd iuainsuisshow studies simulation -body 1 aay ieaisaddnmc aay tutr ehd:nu methods: — structure Galaxy: — dynamics and kinematics Galaxy: astrometry ainlAtooia bevtr fJpn iaa Tokyo Mitaka, Japan, of Observatory Astronomical National AAD1EIEC FDSUTN ESU ARM PERSEUS DISRUPTING OF EVIDENCE DR1 GAIA 2 oiuiMatsunaga Noriyuki , rf eso aur 7 2018 17, January version Draft ABSTRACT mshsRceisiu,Mnhosr1-4 -92 Heid D-69120 12-14, Monchhofstr Recheninstitut, omisches N bd/yrdnmc iuain ae nastatic a on based simulations -body/hydrodynamics n omuyS.Mr,Drig uryR56T UK 6NT, RH5 Surrey Dorking, Mary, St. Holmbury on, 01 rve l 07.Tu,this Thus, for 2017). al. tracer Fernández al. of et also great et (see state a Griv scenario dynamical are 2001; arm they Cepheids the spiral thus the Hence, to testing and sensitive arms. arm, spiral be com- DYN the is to the Cepheids of expected of lifetime are range effects the age dynamical to the to Moreover, parable due eas- exists. motion is it systematic it if a and find dispersion, around to velocity are be ier small have to Cepheids to expected expected are ages 2013). whose 20 al. stars et young (Inno also relation period- ac- well-calibrated are luminosity their distances to their thanks and measured stars curately variable bright are They due it is feature Hence, this arm. arm. not Perseus or Perseus the whether the to conclude from to away difficult far is still is which 21)weetedsacswr eemndhomoge- determined were distances the where (2014) different with simulations the in seen is models. spiral what kinemat- and Cepheids observed results ics the the between presents comparisons 3 of Section properties. kinematic their of state dynamical the arm. study Perseus to the arm Perseus the around nsasi h oa egbuho (distance neighbourhood solar the observed was in torque feature stars the the However, in to arm. owing ro- Perseus high Galactocentric the DYN from whose unexpectedly stars is a of velocity favoring group tation a conclusion finding tentative by arm a reached (2017) Wlsrnewg3,618Hiebr,Germany Heidelberg, 69118 35, -Wolfsbrunnenweg 3 oetJ .Grand J. J. Robert , eew netgt ieaiso ehi variables. Cepheid of kinematics investigate we Here eslce apeo ehisi eoaie al. et Genovali in Cepheids of sample a selected We shows and Cepheids, of sample our describes 2 Section − 0 y Bn ta.20) uhyugsasare stars young Such 2005). al. et (Bono Myr 300 yo oot,OtroMS34 Canada 3H4, M5S Ontario Toronto, of ty o ukok,Tko1303,Japan 113-0033, Tokyo Bunkyo-ku, go, 2. sacso ehisadteproper the and Cepheids of istances EUIRMTOSO CEPHEIDS OF MOTIONS PECULIAR hsi,t u nweg,tefirst the knowledge, our to is, This . i prlseai.Atog our Although scenario. spiral mic saetdb h prlam We arm. spiral the by affected is ocuinta h esu arm Perseus the that conclusion tdJnay1,2018) 16, January pted sami h ik a using Way Milky the in arm us iigside, ailing oprdt h edn side. leading the to compared sacsfo h ou fthe of locus the from istances 8-58 Japan 181-8588, oainvlcte f77 of velocities rotation d 4 , 5 n ao .S Hunt S. A. Jason and , − 27 . 6 ± Letter 2 eia — merical . e,in deg, 4 leg Germany elberg, ssCepheids uses < 0 . kpc) 6 6 2 Baba et al. neously by using near-infrared photometric data sets (also see Inno et al. 2013). Errors in distance modulus are estimated by Genovali et al. (2014) to be 0.05– 0.07 mag for most of the Cepheids. We then cross- matched this sample with the TGAS catalog and with a sample of Cepheids whose radial velocity are provided in Mel’nik et al. (2015) using TOPCAT (Taylor 2005), giv- ing a collection of 206 Cepheids with known locations and kinematics. We further limit the sample based on vertical position with respective to the Sun, |ze,max| < 0.5 kpc, where to take into account the error we define (µdm+µdm,e+5.0)/5.0 ze,max = sin(b)10 /1000 kpc, where b, µdm, and µdm,e are the Galactic latitude, the distance modulus and its error in magnitude, respectively. This limit was applied to eliminate clear outliers, although our sample shows a clear concentration around the Galactic plane, with more than 70 % being located within 100 pc, as expected for young stars like Cepheids. To eliminate the data with a large velocity or distance uncertainty, we discard the data whose uncertainty in velocity, σ = σ2 + σ2 + σ2 , is larger than V q Vlon Vlat VHRV −1 20 km s , where σVlon , σVlat and σVHRV are the uncertainties of the velocity measurements in the direction of longitude, Vlon, latitude, Vlat, and heliocentric radial velocity, VHRV. σVlon and σVlat are computed by taking the Fig. 1.— Face-on distribution of the selected Cepheids with ar- standard deviation of the Monte-Carlo (MC) sampling rows describing their peculiar velocities with color indicating their of V and V , computed for randomly selected right pulsation period (log P ). The cyan, blue, red and black solid lines lon lat show the positions of the Outer, Perseus, Sagittarius and Scutum ascension (RA) and declination (DEC) proper motions, spiral arms, respectively, measured in Reid et al. (2014). The open using the 2D Gaussian probability distribution with their circle indicates the position of the Sun and the arrow shows its pe- measured mean, standard error and correlation between culiar motion. the RA and DEC proper motions, and distance from the Gaussian probability distribution of distance mod- dard of the Rest, (U⊙, V⊙, W⊙) = (10.0 ± 1.0, 11.0 ± −1 ulus with the mean of µdm and a standard deviation of 2.0, 7.0±0.5)km s (Bland-Hawthorn & Gerhard 2016) µdm,e. These selections left 191 Cepheids in our sample. and the radial gradient of circular velocity, dVc/dR = From this sample, we choose Cepheids around the −2.4 ± 1.2km s−1 kpc−1 (Feast & Whitelock 1997). We Perseus arm. We adopt the position of the Perseus arm take the mean and standard deviation of Upec, Vpec and locus as determined by Reid et al. (2014), which provides dPer from the MC sample. the distance and angle to the reference point from the We first look at the correlation coefficients between Sun and the pitch angle of the locus of the arm. We Upec and dPer and between Vpec and dPer. As shown in computed the mean distance, but projected on the Galac- Figures 2a,b, there is a significant positive correlation for tic plane, between the closest point of the locus of the both Upec and Vpec against dPer. We measure the corre- Perseus arm and the Cepheids in our sample, dPer, by lations for each MC sampling described above, and take MC sampling of the distance modulus for Cepheids and the mean and dispersion of the measurements. Table 1 the distance between the Sun and the Galactic center of shows the correlation is statistically significant even after 1 R0 =8.2 ± 0.1 kpc (Bland-Hawthorn & Gerhard 2016) . taking into account the observational errors and uncer- We selected Cepheids within |dPer| < 1.5 kpc, which re- tainties of the Galactic parameters. The correlation is sults in a final catalogue of 77 Cepheids (see Figure 1). stronger in Vpec than Upec. We compute the Galactocentric radial velocity, Upec, Following the idea of Kawata et al. (2014), we com- and rotation velocity, Vpec, after subtracting the circu- pared the velocity distribution of 47 Cepheids on the lar velocity of the disk at the location of each Cepheid. trailing side (defined as 0.2 < dPer < 1.5 kpc) and Again, we used 10,000 MC samples to estimate the un- that of 16 Cepheids on the leading side (−1.5 < dPer < 2 certainties of Upec, Vpec and dPer, taking into account −0.2 kpc) of the Perseus arm . We found a significant the mean and uncertainties of the distance modulus and offset in the mean velocity of these samples as expected proper motion for individual Cepheids and all the rele- from the correlation with dPer (see also Table 1). The vant Galactic parameters, such as R0 = 8.2 ± 0.1 kpc, mean velocity in both hUpeci and hVpeci is higher on the the angular velocity of the Sun with respect to the trailing side. To our knowledge, these results are the first −1 −1 Galactic center, Ω⊙ = 30.24 ± 0.12 km s kpc , the statistically significant observational evidence of the dif- solar peculiar motion with respect to the Local Stan- ference in dynamical properties of stars on different sides of the spiral arm. 1 In our MC sampling, for simplicity we fixed the pitch angle, Furthermore, we calculated the vertex deviation, lv = but only changed the Galactocentric radius of the reference point 2 of the Perseus arm for a sampled R0, although the pitch angle also We excluded Cepheids within |dPer| < 0.2 kpc considering the depends on R0. uncertainty of the locus of the Perseus arm. Perseus Arm Dynamics 3

Fig. 2.— (a) Upec-dPec and (b) Vpec-dPer distributions of our Cepheid sample. Note that Upec is positive in the direction toward the Galactic center, and Vpec is positive in the direction of the Galactic rotation. (c)(d) Upec-Vpec distributions of our Cepheid sample in leading and trailing sides of the Perseus arm, respectively.

TABLE 1 Kinematics of Cepheids and Model results

Cepheids (number) SDW (Rcr = 8 kpc) SDW (Rcr = 16 kpc) DYN (t = 2.59 Gyr) DYN (t = 2.62 Gyr)

Upec − dPer Corr. 0.14 ± 0.02 (77) 0.21 ± 0.07 −0.80 ± 0.02 −0.21 ± 0.03 0.14 ± 0.03 Vpec − dPer Corr. 0.40 ± 0.03 (77) −0.34 ± 0.03 −0.47 ± 0.03 0.062 ± 0.04 0.15 ± 0.03 Trailing side (0.2 < dPer < 1.5 kpc) −1 hUpeci (km s ) 6.1 ± 1.0 (47) −7.0 ± 0.5 −24.3 ± 1.8 −11.4 ± 0.9 4.6 ± 0.8 −1 hVpeci (km s ) −6.3 ± 2.0 (47) −8.5 ± 0.4 −8.5 ± 0.7 −7.5 ± 0.8 −3.8 ± 0.8 lv (deg) −27.6 ± 2.4 (47) −11.5 ± 5.6 3.1 ± 5.1 27.4 ± 5.8 −42.4 ± 5.2 Leading side (−1.5 < dPer < −0.2 kpc) −1 hUpeci (km s ) 0.49 ± 1.2 (16) −10.4 ± 2.9 21.5 ± 0.7 −1.1 ± 1.5 −3.0 ± 1.4 −1 hVpeci (km s ) −13.4 ± 2.0 (16) −1.7 ± 1.1 0.46 ± 0.54 −11.2 ± 1.0 −11.9 ± 0.9 lv (deg) −28.2 ± 6.5 (16) 9.0 ± 3.3 −1.2 ± 10.9 20.8 ± 2.9 22.5 ± 1.9 4 Baba et al.

2 2 2 0.5 × arctan(2σUV /(σU − σV )) (including the correction els. As shown in left side panels of Figure 3, SDW(Rcr = term suggested by Vorobyov & Theis 2008, for the case 16) reproduces none of the observed trend, suggesting 2 of σV > σU ), where σUV is the covariance between Upec that this model is clearly rejected. On the other hand, and Vpec, for the sample on the trailing and leading sides SDW(Rcr = 8) shows some degree of success in the (see their Upec-Vpec distribution in Figures 2c,d). The positive correlation coefficient of Upec-dPer (Figure 3a) results are summarized in Table 1. On the leading side, and a negative vertex deviation in the trailing side (Fig- the number of Cepheids in the sample is too small to ure 3d). However, this model fails to reproduce the posi- measure lv confidently. On the trailing side, in contrast, tive correlation in Vpec-dPer (Figure 3a). Hence, we con- the vertex deviation is clearly negative, which is opposite clude that irrespective of Rcr (i.e., the pattern speed), to the positive one (about +20 deg) of the young stars it is difficult for the SDW models to explain the ob- in the local solar neighborhood (e.g. Dehnen & Binney served features in Section 2. This does not mean that 1998; Rocha-Pinto et al. 2004). To our knowledge, this we can reject the SDW scenario. Indeed, we have not is the first detection of the change of sign of the vertex explored models of different pitch angles and/or different deviation near the spiral arms. Such a change is expected strength of the arms; moreover, our SDW models do not in various spiral arm models (Roca-Fàbrega et al. 2014). include the bar, which is likely to affect the dynamical Finally, we discuss the effect of interstellar redden- features (e.g. Monari et al. 2016). Although these dif- ing and extinction considering its importance for the ferent kinds of models need to be tested against our ob- objects in the disk (Matsunaga 2017). Inno et al. served Cepheid kinematics, the SDW spiral models tend (2013) assumed the total-to-selective reddening ra- to show a regular trend in stellar kinematics around the tio of Cardelli et al. (1989), which is different from spiral arms (e.g. Roca-Fàbrega et al. 2014; Pasetto et al. some of the recent values (e.g. Nishiyama et al. 2006; 2016; Antoja et al. 2016). We therefore expect that the Alonso-Garc´ıaet al. 2017). The reddenings of the positive correlation of Vpec-dPer is difficult to obtain in Cepheids around the Perseus arm are, however, relatively the SDW model alone. small, EJ−Ks ≤ 0.5 mag, and the uncertainty caused by We then compare the DYN model with the observa- the extinction law (up to 7 % in distance) does not change tions. The right side panels of Figure 3 show the re- our results. sults around a spiral arm which grew and was disrupted around t =2.59 and t =2.62 Gyr (indicated with vertical 3. DYNAMICAL NATURE OF THE PERSEUS ARM dot-dashed lines), respectively. As shown in Baba et al. We compare our findings with N-body/hydrodynamic (2013) and Grand et al. (2014), the kinematic properties simulations with different spiral models. The sim- of the stars around the DYN spiral arms change with ulations include self-gravity, radiative cooling, star time. As a result, the growing phase of the DYN arm formation, and stellar feedback (Saitoh et al. 2008; (at t ∼ 2.59 Gyr) is not consistent with the observed Saitoh & Makino 2009). The DYN arm model is a barred properties. spiral galaxy formed from an initial axisymmetric model, Among our models, the disruption phase (at t ∼ 2.62 spontaneously. The bar is an almost stable pattern, but Gyr) of the DYN arm is qualitatively the best at repro- the amplitudes, pitch angles, and rotational frequencies ducing the observed trends. The correlation coefficients of the spiral arms change within a few hundred mil- between velocity are both positive as observed, although lion years (Baba 2015). The SDW models are from the correlation is stronger for Vpec-dPer in the observa- Baba et al. (2016) and have a rigidly rotating two-armed tional data (Figure 3a). As shown in Figures 3b and spiral (external potential) with a pitch angle of 12 deg 3c, both hUpeci and hVpeci are also in good agreement and a spiral amplitude of 3 %. To study the impact with our observational results, and the trailing side shows of the location of the co-rotation radius (Rcr), we used higher values than the leading side. The observed neg- two SDW models with Rcr = 8 kpc (e.g. Fernández et al. ative vertex deviation is also reproduced in the trailing 2001) and 16 kpc (e.g. Lin et al. 1969). side. On the other hand, the leading side shows less To compare the simulations with the observational sensitivity of vertex deviation to the phase of the DYN data, we have applied the same analysis as Section 2 arm, i.e. always positive, which is inconsistent with our for the simulations. First we identify a spiral arm sim- observed trend (Figure 3d). However, the measurement ilar to the Perseus arm in terms of the Galactocentric of vertex deviation in the leading side is less reliable. radial range. Then, we selected young star particles (50– Hence, the disruption phase of the DYN arm shows qual- 200 Myr) around the arm which are located in a radial itative agreement with the high-confidence results of our and azimuthal range similar to that of our Cepheids sam- Cepheid data. Considering that our N-body simulations ple. Note that the pitch angle of our SDW models are are still far from the real Milky Way because of lack of not tuned to match the Perseus arm, but the one to best physical processes and lack of observational constraints, explain both the Scutum and Perseus arms with a sin- it is striking to find this level of agreement between the gle pitch angle. Also, the pitch angle of the DYN model disruption phase of our simulated DYN arm and the ob- is changing as time goes on. Thus, for both SDW and served trends found in our Cepheid data. We therefore DYN models we measure the distance of the particles conclude that the disruption phase of a DYN spiral arm from the gas arm (to be consistent with the identifica- like seen in our simulation, is the most likely scenario for tion of the arm by the star forming regions in Reid et al. the Perseus arm in the Milky Way. 2014), darm, irrespective of the pitch angles of the spiral It is known that the age of Cepheids is well corre- arms, and consider it same as dPer for our Cepheids data lated with their pulsation period (log P ) (e.g. Bono et al. analysis. The results of the simulations are summarized 2005). We color coded Cepheids by log P in Figure 1. We in Table 1, and are shown in Figure 3. found no clear correlation between the age of Cepheids We first compare the observation with the SDW mod- with the position with respect to the arm. The SDW sce- Perseus Arm Dynamics 5 nario predicts a clear correlation between the age of stars by the Japan Society for the Promotion of Science and the distance from the arm (e.g. Dobbs & Pringle (JSPS) Grant-in-Aid for Young Scientists (B) Grant 2010). Hence, this is also against the prediction from Number 26800099. DK acknowledges the support of the the SDW scenario, but more consistent with the DYN UK’s Science & Technology Facilities Council (STFC arm scenario (e.g. Grand et al. 2012a). Grant ST/N000811/1). NM is grateful to Grant-in-Aid Interestingly, according to Reid et al. (2014), the pitch (KAKENHI, No. 26287028) from the Japan Society for angle of the Perseus arm (9.4 ± 1.4 deg) is smaller than the Promotion of Science (JSPS). JASH is supported the Scutum arm (19.8±2.6 deg), which is the other major by a Dunlap Fellowship at the Dunlap Institute for As- arm. N-body simulations of DYN arms predict that the tronomy & Astrophysics, funded through an endowment pitch angle of spiral arms in the disruption phase would established by the Dunlap family and the University be smaller, because the arms are winding and disrupt- of Toronto. RG acknowledges support by the DFG ing (Baba et al. 2013; Grand et al. 2013). Therefore, if Research Centre SFB-881 ‘The Milky Way System’ the Perseus arm does indeed have a small pitch angle, it through project A1. This work has made use of data is also consistent with the arm being in the disruption from the European Space Agency (ESA) mission Gaia phase. We will further test the disruption phase scenario (https://www.cosmos.esa.int/gaia), processed by the of the Perseus arm with the future Gaia data releases. Gaia Data Processing and Analysis Consortium (DPAC, https://www.cosmos.esa.int/web/gaia/dpac/consortium). Funding for the DPAC has been provided by national We are grateful to the referee for useful suggestions institutions, in particular the institutions participating that helped improve this manuscript. We also thank in the Gaia Multilateral Agreement. The simulations Nobuyuki Sakai for much useful advice on analysis reported in this paper were carried out on facilities of astrometric data. We thank Jo Bovy for making of Center for Computational Astrophysics (CfCA), galpy (Bovy 2015), which is used for coordinate National Astronomical Observatory of Japan. transformation, publicly available. JB was supported

REFERENCES Alonso-Garc´ıa, J. et al. 2017, ApJ, 849, L13 Hunt, J. A. S., Kawata, D., Grand, R. J. J., Minchev, I., Pasetto, Antoja, T., Roca-Fàbrega, S., de Bruijne, J., & Prusti, T. 2016, S., & Cropper, M. 2015, MNRAS, 450, 2132 A&A, 589, A13 Hunt, J. A. S., Kawata, D., Monari, G., Grand, R. J. J., Famaey, Baba, J. 2015, MNRAS, 454, 2954 B., & Siebert, A. 2017, MNRAS, 467, L21 Baba, J., Asaki, Y., Makino, J., Miyoshi, M., Saitoh, T. R., & Inno, L. et al. 2013, ApJ, 764, 84 Wada, K. 2009, ApJ, 706, 471 Kawata, D., Hunt, J. A. S., Grand, R. J. J., Pasetto, S., & Baba, J., Morokuma-Matsui, K., Miyamoto, Y., Egusa, F., & Cropper, M. 2014, MNRAS, 443, 2757 Kuno, N. 2016, MNRAS, 460, 2472 Lin, C. C., & Shu, F. H. 1964, ApJ, 140, 646 Baba, J., Saitoh, T. R., & Wada, K. 2013, ApJ, 763, 46 Lin, C. C., Yuan, C., & Shu, F. H. 1969, ApJ, 155, 721 Bertin, G., & Lin, C. C. 1996, Spiral structure in galaxies a Lindegren, L. et al. 2016, A&A, 595, A4 density wave theory, ed. G. Bertin & C. C. Lin Matsunaga, N. 2017, in European Physical Journal Web of Bland-Hawthorn, J., & Gerhard, O. 2016, ARA&A, 54, 529 Conferences, Vol. 152, European Physical Journal Web of Bono, G., Marconi, M., Cassisi, S., Caputo, F., Gieren, W., & Conferences, 01007 Pietrzynski, G. 2005, ApJ, 621, 966 Mel’nik, A. M., Rautiainen, P., Berdnikov, L. N., Dambis, A. K., Bovy, J. 2015, ApJS, 216, 29 & Rastorguev, A. S. 2015, Astronomische Nachrichten, 336, 70 Cardelli, J. A., Clayton, G. C., & Mathis, J. S. 1989, ApJ, 345, Michalik, D., Lindegren, L., & Hobbs, D. 2015, A&A, 574, A115 245 Monari, G., Famaey, B., Siebert, A., Grand, R. J. J., Kawata, D., Dehnen, W., & Binney, J. J. 1998, MNRAS, 298, 387 & Boily, C. 2016, MNRAS, 461, 3835 Dobbs, C., & Baba, J. 2014, PASA, 31, 35 Nishiyama, S. et al. 2006, ApJ, 638, 839 Dobbs, C. L., & Pringle, J. E. 2010, MNRAS, 409, 396 Pasetto, S., Natale, G., Kawata, D., Chiosi, C., Hunt, J. A. S., & D’Onghia, E., Vogelsberger, M., & Hernquist, L. 2013, ApJ, 766, Brogliato, C. 2016, MNRAS, 461, 2383 34 Reid, M. J. et al. 2014, ApJ, 783, 130 Feast, M., & Whitelock, P. 1997, MNRAS, 291, 683 Roca-Fàbrega, S., Antoja, T., Figueras, F., Valenzuela, O., Fernández, D., Figueras, F., & Torra, J. 2001, A&A, 372, 833 Romero-Gómez, M., & Pichardo, B. 2014, MNRAS, 440, 1950 Fujii, M. S., Baba, J., Saitoh, T. R., Makino, J., Kokubo, E., & Rocha-Pinto, H. J., Flynn, C., Scalo, J., Hänninen, J., Maciel, Wada, K. 2011, ApJ, 730, 109 W. J., & Hensler, G. 2004, A&A, 423, 517 Gaia Collaboration et al. 2016a, A&A, 595, A2 Saitoh, T. R., Daisaka, H., Kokubo, E., Makino, J., Okamoto, T., —. 2016b, A&A, 595, A1 Tomisaka, K., Wada, K., & Yoshida, N. 2008, PASJ, 60, 667 Genovali, K. et al. 2014, A&A, 566, A37 Saitoh, T. R., & Makino, J. 2009, ApJ, 697, L99 Grand, R. J. J., Kawata, D., & Cropper, M. 2012a, MNRAS, 426, Sellwood, J. A., & Carlberg, R. G. 1984, ApJ, 282, 61 167 Taylor, M. B. 2005, in Astronomical Society of the Pacific —. 2012b, MNRAS, 421, 1529 Conference Series, Vol. 347, Astronomical Data Analysis —. 2013, A&A, 553, A77 Software and Systems XIV, ed. P. Shopbell, M. Britton, & —. 2014, MNRAS, 439, 623 R. Ebert, 29 Grand, R. J. J. et al. 2016, MNRAS, 460, L94 Vorobyov, E. I., & Theis, C. 2008, MNRAS, 383, 817 Griv, E., Hou, L.-G., Jiang, I.-G., & Ngeow, C.-C. 2017, MNRAS, 464, 4495 6 Baba et al.

Fig. 3.— Comparison between the models and the observed kinematics of Cepheids in terms of correlations of Upec − dPer and Vpec−dPer (panel a), mean Upec (panel b), mean Vpec (panel c) and the vertex deviation, lv (panel d). Horizontal shaded areas in black and red indicate the observed values and the 1σ uncertainty ranges for Upec −dPer and Vpec −dPer) correlation, respectively, in the top panel, while the measured values in the trailing (leading) side of the arm are indicated by diﬀerent colors in the other panels. The model results are shown with the open symbols with error bars. Black stars (red square) in the top panel shows the Upec − dPer (Vpec − dPer) correlation. In the second, third and bottom panels, black square (red triangle) show the measured values in the trailing (leading) side of the arm. Note that the left side of the panels shows the SDW model results for two diﬀerent Rcr values, whereas in the right side the DYN model results are presented as a function of time. These results are also summarized in Table 1.