Exploring the Galactic Anticenter substructure with LAMOST & Gaia DR2
Jing Li, Xiang-Xiang Xue, Chao Liu, Bo Zhang, Hans-Walter Rix, Jeffrey L. Carlin, Chengqun Yang, Rene A. Mendez, Jing Zhong, Hao Tian, Lan Zhang, Yan Xu, Yaqian Wu, Gang Zhao, Ruixiang Chang
DDraft version January 8, 2021
Typeset using L A TEX default style in AASTeX62
Exploring the Galactic Anticenter substructure with LAMOST & Gaia DR2
Jing Li,
1, 2
Xiang-Xiang Xue,
3, 4
Chao Liu,
3, 4
Bo Zhang,
3, 4
Hans-Walter Rix, Jeffrey L. Carlin, Chengqun Yang, Rene A. Mendez, Jing Zhong, Hao Tian, Lan Zhang, Yan Xu, Yaqian Wu, Gang Zhao,
3, 4 and Ruixiang Chang Physics and Space Science College, China West Normal University, 1 ShiDa Road, Nanchong 637002, P.R.China Chinese Academy of Sciences South America Center for Astronomy, National Astronomical Observatories, CAS, Beijing 100012, China CAS Key Laboratory of Optical Astronomy, National Astronomical Observatories, Chinese Academy of Sciences, Beijing 100101, China School of Astronomy and Space Science, University of Chinese Academy of Sciences, 19A Yuquan Road, Shijingshan District, Beijing100049, China Max-Planck-Institute for Astronomy K¨onigstuhl 17, D-69117, Heidelberg, Germany AURA/Rubin Observatory, 950 North Cherry Avenue, Tucson, AZ 85719, USA Shanghai Astronomical Observatory, Chinese Academy of Sciences, 80 Nandan Road, Shanghai 200030, China Departamento de Astronomia, Universidad de Chile, Casilla 36-D, Correo Central, Santiago, Chile Key Laboratory for Research in Galaxies and Cosmology, Shanghai Astronomical Observatory, Chinese Academy of Sciences, 80 NandanRoad, Shanghai 200030, China (Received xxx; Revised xxx; Accepted xxx)
Submitted to ApJABSTRACTWe characterize the kinematic and chemical properties of 589 Galactic Anticenter SubstructureStars (GASS) with K-/M- giants in Integrals-of-Motion space. These stars likely include membersof previously identified substructures such as Monoceros, A13, and the Triangulum-Andromeda cloud(TriAnd). We show that these stars are on nearly circular orbits on both sides of the Galactic plane.
We can see velocity( V Z ) gradient along Y-axis especially for the south GASS members. Our GASS members have similar energy and angular momentum distributions to thin disk stars. Theirlocation in [ α /M] vs. [M/H] space is more metal poor than typical thin disk stars, with [ α /M] lower than the thick disk. We infer that our GASS members are part of the outer metal-poor disk stars, andthe outer-disk extends to 30 kpc. Considering the distance range and α -abundance features, GASScould be formed after the thick disk was formed due to the molecular cloud density decreased in theouter disk where the SFR might be less efficient than the inner disk. Keywords: galaxy: individual (Milky Way) – Galaxy: halo – Galaxies: structure – stars: M giantstars – stars: kinematics and dynamics INTRODUCTION1.1.
Historical Overview of Galactic Anticenter substructure stars
The low Galactic latitude substructures collectively known as “The Monoceros Ring” (thereafter referred to as Mon)were first discovered by Newberg et al. (2002). Four pieces of stellar over-densities were identified at low latitude– those labeled S223+20-19.4, S218+22-19.5, and S183+22-19.4 located in the north Galactic cap, and S200-24-19.8found below the plane. All of these detections are within 40 ◦ of the Galactic anticenter at ( l, b ) = (180 ◦ , ◦ ). Theturnoffs of the faint main sequence of these overdensities are significantly bluer than the turnoff of typical thick diskstars. Newberg et al. (2002) interpreted the bluer turnoff as evidence for a more metal-poor population, as had been Corresponding author: Xiang-Xiang Xue; Jing [email protected]; [email protected] a r X i v : . [ a s t r o - ph . GA ] J a n Li et al. identified in the Sagittarius dwarf tidal stream, though the effect could also be due to a population with a youngerstellar age.Within a year, Yanny et al. (2003) and Ibata et al. (2003) suggested that the low latitude structure was ring-like,and could potentially encircle the entire galaxy (this same structure has been referred to as the Galactic AnticenterStellar Structure, or GASS, by Rocha-Pinto et al. 2003 and subsequent work). Yanny et al. (2003) traced the structurefrom 180 ◦ < l < ◦ , and noted that it extended 5 kpc above and below the Galactic plane, though the southernportion was 2 kpc further away. They found a velocity dispersion of ∼
25 km s − , which is lower than that of thethick disk (typically 40-50 km s − ), and a scale height that is larger than the thick disk. Based on these observations,Yanny et al. (2003) argued that“the Monoceros ring” is a tidal stream. Simultaneously, Ibata et al. (2003) pointedout that the ring was not actually traced around the entire Galaxy, but was detected in the range 120 ◦ < l < ◦ ,on both sides of the Galactic plane. This suggests that the ring could be an artifact in the disk caused by repeatedwarping, a tidal stream from an accreted satellite, or part of an outer spiral arm. In the same year, Rocha-Pinto etal. (2003) showed that the ring was visible in 2MASS M giant stars, establishing that there is a range of metallicitiespresent in this structure, and believed it was the result of a tidally disrupting dwarf galaxy (for more details aboutMonoceros/GASS, see Yanny & Newberg 2016 and other chapters in Newberg & Carlin 2016).Several authors have raised the hypothesis of a flare and warp of the outer disk to explain the Monoceros Ring(Momany et al. 2006; Moitinho et al. 2006; Carraro & Costa 2009; Hammersley & L´opez-Corredoira 2011; L´opez-Corredoira et al. 2002; Feast et al. 2014). However, this idea has difficulties explaining the narrow radial velocitydispersion of Mon (Meisner et al. 2012).An extensive photometric and spectroscopic study using SDSS, including SEGUE and SEGUE-2 spectroscopy in theanticenter region, was carried out by Li et al. (2012). They conclude that the Monoceros structure has a metallicity of[Fe/H] ∼ − . ± .
1, and the ring has a higher metallicity than the halo, but slightly lower metallicity and a narrowervelocity dispersion than the thick disk. All kinematics of Monoceros stars, however, show prograde motion, rotatingwith the disk, and not far in velocity from the circular motion of stars in a flat rotation curve.Subsequently, the Pan-STARRS1 survey presented a panoramic picture of the anticenter in stellar density above andbelow the Milky Way plane. Figure 2 of Slater et al. (2014) clearly shows the “band-like” structure stretching from l = 100 ◦ to 230 ◦ , and covering a large Galactic latitude range of − ◦ < b < ◦ in some regions.Detailed studies of chemical abundances in Mon/GASS have been carried out by Chou et al. (2010) and Meisneret al. (2012). Chou et al. (2010) derived chemical abundance patterns from high-resolution spectra of 21 M giants.The abundances of the α -element titanium, and s-process elements yttrium and lanthanum, for these GASS stars arefound to be lower at the same [Fe/H] than those for MW stars, but similar to those of stars in the Sagittarius stream,other dwarf spheroidal galaxies, and the Large Magellanic Cloud. From low resolution ( R ∼ ∼ −
1, which is intermediate between the halo and the local thick disk. They also found significantly differentmetallicities for stars at the same Galactocentric radii above and below the plane ([Fe/H]= − .
65 vs. − .
87 dex,respectively).Xu et al. (2015) showed that there is a vertical asymmetry in the disk that is a function of distance from the Galacticcenter, as measured in the direction of the anticenter. They think the oscillation lines up with the position and densityof the Mon and TriAnd structures, but it is only apparent in the north. There are excess star counts in both the northand the south, but they are at different Galactocentric distances.Li et al. (2017) presented an analysis of spectroscopic observations of individual stars from “A13”, which theyconcluded based on positions, distances, and kinematical properties to be an extension of Mon. Sheffield et al. (2018)studied the stellar population of Mon, A13, and TriAnd, to assess the relative numbers of RR Lyrae and M giantstars, and found that both structures have very low f RR : MG , supporting the scenario in which stars in Mon, A13, andTriAnd formed in the MW disk.1.2. Historical overview of MWTD in the context of the Milky Way
The metal-weak thick-disk(MWTD) population has been confirmed existence over the past twodecades. The first paper about MWTD was from Norris et al. (1985), which presented a sampleof 309 non-kinematically weak-metal candidates in the solar neighborhood. Follow-up research sug-gested that the low-metallicity and low-eccentricity stars belong to a population that is intermediate in its motion perpendicular to the Galactic plane between that of the thin disk and that of metal-deficientobjects of extreme eccentricity and the velocity dispersion of this group stars is consistent with thickdisk(Morrison et al. 1990; Beers et al. 1995).The other observational efforts include high/ medium-resolution spectroscopic abundance determina-tions have addressed the problem(the fractions of stars at low metallicity were substantially smaller) ofthe existence of a MWTD component in the Solar Neighborhood of Galaxy (Beers & Sommer-Larsen1995; Chiba & Yoshii 1998; Martin & Morrison 1998; Beers et al. 2000; Arifyanto et al. 2005). Someauthors interpreted the origin of the MWTD in terms of the debris of a ”shredded satellite”(Gilmoreet al. 2002; Arifyanto et al. 2005). The preponderance of evidence acquired prior to 2009 suggestedthat a MWTD component exists in the Solar Neighborbood, although these analyses generally did notconsider stars with such a large lag as likely candidate disk-like stars. As argued by Villalobos & Helmi(2009).In the period since 2009, a substantial volume of work has been carried out, making use of large sam-ples of stars with medium-resolution(R ∼ ∼ − . < [ F e/H ] < − . , however, notrotationally supported. Beers et al. (2014) using a new set of very high signal-to-noise(S/N > R ∼ ) optical spectra obtained for 302 of the candidate ”weak-metal” stars se-lected by Bidelman & MacConnell (1973), this work proved the presence of MWTD population, andfound 25 % of the stars with metallicities − . < [ F e/H ] < − . exhibit orbital eccentricites e < . ,yet are clearly separated from members of the inner-halo population with similar metallicities. Theseworks have raised new and interesting questions concerning the nature of the formation and evolutionof both the disk and halo systems which need a large spectroscopic data to arrive at a widely acceptedview. The second data release of the Gaia mission (Gaia Collaboration et al. 2018), in combination with large sample ofspectroscopic surveys, provides the opportunity to search the Galactic halo substructures in a wide view in 7D phasespace (i.e., 6D positions+velocities, plus metallicities). Xue et al. (in preparation) identified substructures with highreliability in energy versus angular momentum space with K/M-giants selected from the LAMOST spectroscopic surveyand Gaia DR2. From these large groups of substructures we found 4 ring-like groups around the Galactic anticenter,with kinematic features similar to the GASS. In this paper, we present the kinematic and chemical features of thesegroups. The paper is organized as follows. In Section 2 we describe LAMOST K/M giant stars and the Integralsof Motion and Friends-of-Friends algorithm. The kinematic features of GASS stars are described in Section 3. Wepresent the chemical abundance features in Section 4. The discussion and conclusion are presented in Section 5. DATA AND METHOD2.1.
Data
The data used in this work consist of spectroscopically-identified K- and M-giants from LAMOST Data Release 5(DR5). The LAMOST Telescope is a 4m Schmidt telescope placed at Xinglong Observing Station. This NationalKey Scientific facility built by the Chinese Academy of Sciences (Cui et al. 2012; Zhao et al. 2012; Luo et al. 2012;Deng et al. 2012) has finished the first stage of its regular survey (LAMOST-I, from 2011-2017; including the pilotsurvey), and provided 9,027,634 low-resolution (R ∼ , K-giant stars are selected from LAMOST DR5 according to the criteria presented in Liu et al.2014 ( (4000K < T eff < < . (cid:107) (4600K < T eff < < ). The distances of K-giants areestimated following the Bayesian method described in Xue et al. (2014), which is suitable for distance Li et al. estimation of halo K-giants due to the adopted fiducials of globular clusters. The typical distanceprecision of LAMOST halo K-giants is 13% (Yang et al. 2019; Bird et al. 2019).
M-giant stars are from Zhong & Li et al. (2019), which used a spectroscopic template matching method plus2MASS+WISE photometric selection to identify 40,000 M-giants from LAMOST DR5. The contamination of theM-giant sample by 894 carbon stars have also been excluded by cross-matching with the latest LAMOST carbon starcatalog (Ji et al. 2016; Li et al. 2018). The distances of M-giants were calculated through the ( J − K ) color distancerelation derived by Li et al. (2016). Since the distance of K- and M-giants are derived from different calculation method and calibrationwith distant stars. We re-calibrated the distances of K- and M-giants with Gaia DR2 parallax ratherthan Gaia distances estimated by Bailer-Jones et al. (2018). Because Bailer-Jones et al. (2018) claimedthat their mean distances to distant giants are underestimated, due to the stars have very large frac-tional parallax uncertainties, so their distance are prior-dominated, and the prior was dominated bythe nearer dwarfs in the model. Only stars with good parallaxes ( δ(cid:36)/(cid:36) < ) and good distances( δd/d < ) are used to do the calibration, which allows us to compare parallax with /d , and min-imize the possible bias from inverting. Finally, we used halo stars(the sample which deleted all theidentified group members) selected out from Xue 2019(in prepare) as the calibration sample. For K-and M-giants there are and bias respectively( detail see in figure 1 of Yang et al. (2019) ).Therefore, we increase the distances of K-giants by (1 / . − ∼ and decreased the distance ofM-giants by . The proper motions of the K- and M-giants are obtained by crossing match with Gaia DR2 less than 1 (cid:48)(cid:48) . LAMOSTpipeline provides the heliocentric radial velocities hrv with the typical error of 7 km − . The chemical abundance(metallicity [M/H] and abundance of α -element [ α /M]) of LAMOST K- and M-giants are from Zhang et al. (2019),which used a machine learning program called Stellar LAbel Machine (SLAM) to transfer APOGEE (Majewski etal. 2017) stellar labels to LAMOST DR5 spectra. The corresponding cross-validated scatters of [M/H] and [ α /M] athigh SNR g ( ∼ All stars for K- and M-giants did the initial cut, only starsat | Z | > kpc and those with 2 kpc < | Z | < < − are classified as halo stars whichwe will use to identify substructures in the following section. The disk stars we used in this work ascomparison with our GASS members are selected from LAMOST K-giants which | Z | < kpc (detail insection 3.3). Identification of Galactic Anti-center Substructure
Substructure in the Galaxy can be taken as stars moving on similar orbits, but possibly on quite different orbitalphases. The orbit can be characterized by its integrals of motion (
I.o.M ). Under the assumption that the potential ofGalactic halo, in the simplest approximation, is relatively spherical, there are four integrals of the motion: theenergy E , the angular momentum vector (cid:126)L ( (cid:126)L x , (cid:126)L y and (cid:126)L z ) . As described in Yang et al. (2019), Xue et al. inpreparation identified Galactic substructure through grouping stars with similar E and (cid:126)L from LAMOST K-/M-giants,SEGUE K-giants and SDSS BHBs. Specifically, the E and (cid:126)L can be translated to eccentricity e , semi-major axis a ,the direction of the orbital pole ( l, b ) orbit and the direction of apocenter l apo (i.e. the angle between apocenter and theprojection of x-axis on the orbital plane). Please note that l apo changes with periods, but keeps constant within oneperiod, which can be used to distinguish stars in the same stream but involving in our Galaxy in different epochs (e.g.Sgr leading and trailing arms).By defining the orbit-likelihood-distance to measure how close two stars distribute in I.o.M. space, Xue et al. inpreparation applied friends-of-friends (FoF) algorithm to link stars moving on similar orbits together, where the choiceof ”linking length” is to make sure Sgr streams can be identified as complete as possible.
There are four data set used in Xue et al. work to identify halo substructures, include LAMOSTK-giants and M-giants, SEGUE K-giants and SDSS BHBs.
Finally, Xue et al. in preparation identified27 groups from LAMOST K-/M-giants and SDSS K-giants and BHB stars with group members larger than 50. Bycomparing the sky coverage of these groups, we find four groups from LAMOST K-/M-giants located in Galacticanti-center which share the same kinematic and chemical properties. Two of these four groups are from the K-giantssample, and the others from the M-giants sample. Their location in Galactic coordinates can be seen in Figure 1.
These four groups share similar kinematic and chemical features on the two sides of the Galactic disk(in the following section, we will discuss the kinematic and chemical features in detail). If we relax the restriction of link-length from the FoF method in Xue et al. (in preparation), the two groups ofK-giants will merge into one group because they located in two sides of Galactic disk, this condition issame to two groups of M-giants. Considering these four groups also share same kinematic and chemicalproperties, we infer that they belong to the same structure.
By comparing these candidates with results fromvarious works (Newberg et al. 2002; Ibata et al. 2003; Yanny et al. 2003; Li et al. 2012; Slater et al. 2014), we find thatour candidates actually include at least the three components Mon, A13, and Triand. Mon is the ring-like substructuredetected in the lower latitudes near the Galactic anticenter, with line of sight velocity similar to the thick disk butmuch smaller velocity dispersion (Newberg et al. 2002; Yanny et al. 2003; Li et al. 2012). A13 is a substructure foundto the north of the Galactic plane in the anticenter direction, with heliocentric distance ∼ −
20 kpc, line of sightvelocity distribution similar to the disk, and [
F e/H ] ∼ − . − . ∼ −
25 kpc,line of sight velocity distribution similar to a disk model with rotation velocity 150 km s − , and distance ∼
20 kpc(Deason et al. 2014). The truth is that there are many substructures in the Galactic anticenter region, and many ofthem overlap in their properties (e.g., spatial distribution, kinematics, and chemical abundances). Our work collectsthem all as one group with K-/M-giants.Throughout this work, we refer to the collection of these structures as the Galactic Anticenter Substructure Stars(GASS thereafter), with the caveat that the features may not all be part of the same structure. In the next section, wewill exhibit the features of these candidates in detail, including spatial, kinematic and chemical abundance features.The observational parameters and calculated orbit parameters of the 589 GASS members are listed in Table 1 and 2separately. In what follows, we will further explore the nature of GASS using this sample. THE KINEMATIC FEATURES OF GASS3.1.
Spatial distribution
Figure 1 summarizes the Galactic distribution of the selected GASS members. The members cover a large area ontwo sides of the disk near the Galactic anticenter, spanning the range 90 ◦ < l < ◦ and − ◦ < b < ◦ . We furthercompare the positions of these stars with the density map of main-sequence turnoff stars from the work of Slater et al.(2014) based on the Pan-STARRS catalog. Figure 1 is similar to Figure 3 from Slater et al. (2014), but using coloreddots that label stars at different distance. Our GASS members have good positional alignment with the ”band-like”structures in the Pan-STARRS map. It is worth noting that our sample is in good agreement with the GASS featureshighlighted in their paper (see, e.g., features B, C, and D in Figure 3 of Slater et al. (2014)). We also labeled outprevious detected Mon,A13 and TriAnd regions with white, green and purple squares separately ascomparison.
The Phase-space Distribution
For the measurement errors of our sample, LAMOST K-giants have a median distance precision of 13% (Xue etal. 2014), a median radial velocity error of 7 km s − , a median error of 0.14 dex in metallicity, and a median [ α /Fe]error of 0.05 dex (Liu et al. 2014; Zhang et al. 2019). LAMOST M-giants have a typical distance precision of 20%,but do not have estimates of the distance error for each star (Li et al. 2016). LAMOST M-giants have typical radialvelocity errors of about 5 km s − (Zhong & Li et al. 2019), a median error of 0.17 dex in metallicity, and a median[ α /Fe] error of 0.06 dex (Zhang et al. 2019). The proper motions of K-giants and M-giants are derived from Gaia
DR2, which is good to 0.2 mas yr − at G=17 m . In this work, we calculate the errors of each calculated parameterfor all K-/M-giants, based on the observational parameters (such as rv, pmra, pmdec, heliocentric distance) using anMCMC method. This involves running our algorithm 1000 times for each single star, sampling from the parametererror distributions (assumed Gaussian) to get per-star errors for the different parameters. With these parameters andcorresponding errors, we are able to analyze the phase-space distribution of the 589 GASS members; the parametersand corresponding errors can be found in Table 1 and Table 2. Xu et al. (2015) proposed that Mon (close to theSun) and TriAnd (farther from the Sun) could be associated with the same locally apparent disturbance, as thenorthern and southern parts of a vertically oscillating ring propagating outward from the Galactic center. G´omez etal. (2016) and Laporte et al. (2018) using N-body and/or hydro-dynamical simulations have shown that Milky Waysatellites could produce strong disturbances and might lead to the formation of vertical structure in the Galactic disk.Some observational work has also shown that there exists a ripple pattern and perturbed velocities in the disk within r gc <
12 kpc (Liu et al. 2017; Antoja et al. 2018; Tian et al. 2018; Binney & Sch¨onrich 2018; Cheng et al. 2019;
Li et al. -60° -30° 0°+30°+60°
MonA TriAnd
K-giantsM-giants
B C D
Figure 1. Sky coverage of our GASS members in Galactic coordinates. The blue and red filled circles showour GASS Members with K-/M-giants. The background shows sky coverage map of main sequence turn offstars from the Pan-STARRS catalog with . < ( g − r ) < . . Nearby stars with . < g < . . − . kpc) are shown in blue, stars with . < g < . . − . kpc) are shown in green, and more distant starswith . < g < . . − . kpc) are shown in red. The green/white/purple square regions show previousdetected A13, Mon and TriAnd regions. Bland-Hawthorn et al. 2019; Laporte et al. 2019). Figure 8 from Li et al. (2017) schematically illustrates a possiblescenario where Mon, A13, and TriAnd are the signatures of disk oscillations at different Galactocentric distances.Figure 2 shows projections of the 3D distribution of our GASS samples onto the Galactic X − Y and r − Z planes .In the left panel, the arrows indicate the direction and amplitude of velocities in the X − Y plane. We can clearlysee our GASS have circle orbit in X − Y plane. In the right panel, we can see that this structure covers a largerange in r gc , from 15 kpc out to 30 kpc. This range actually overlaps with the distances and part of sky areas to Mon,TriAnd, and A13, since there are not clear boundaries between all of these features. The blue and red error bars in thefigure show the mean errors for X, Y, and r gc . Because M-giants do not have distance errors for each star, we assigna distance error for M-giants with a random function with restriction of relative error equal 0.05. Figure 3 show spatial distribution of our GASS in the Y-Z plane. In the upper left panel, the arrowsrepresent all GASS members moving direction. We can see a little arc shape from the motion especiallyin the south hemisphere.
In the remaining three panels, we color-code regions by the mean V Z component of thestars’ velocities for the samples above/below the plane. From the upper right panel, we see that combined K-/M-giantsGASS members show clear V Z gradient along the Y axis . The systematical distance error in this work can beignored because we did calibration for K-giants and M-giants with Gaia parallax. We also make similar plots forK-giants and M-giants separately in the lower two panels. From these two sub-samples we can also see a clear V Z gradient in the Y direction, especially in the M-giants sample, confirming our finding from the combined samples. Thisbehavior could be part of expected ripple pattern extending out to r gc >
15 kpc by Xu et al. (2015) and Li et al.(2017).
Figure 4 shows the line-of-sight velocity distribution for K- and M-giants members of GASS in northand south hemisphere separately. In the upper panel, shows the north part K- and M-giants groupsline-of-sight velocity distribution, the mean velocities for K- and M-giants groups are -20.33 kms − and-26.87 kms − , the velocity dispersion are 44.08 kms − and 32.32 kms − . In the lower panel, shows thesouth part K- and M-giants groups line-of-sight velocity distribution, the mean velocities for K- andM-giants are 17.91 kms − and 16.27 kms − , the velocity dispersion are 17.31 kms − and 28.86 kms − . The The Cartesian reference frame used in this work is centered at the Galactic center, the X -axis is positive toward the Galactic center,the Y -axis is along the rotation of the disk, and the Z -axis points toward the North Galactic Pole. The Sun’s position is at (-8.3,0,0)kpc (de Grijs & Bono 2016), the local standard of rest (LSR) velocity is 225 km s − (de Grijs & Bono 2017), and the solar motion is(+11 . , +12 . , +7 .
25) km s − (Sch¨onrich et al. 2010). typical line-of-sight velocity dispersion for thick disk is around 30-40 kms − (Bensby et al. 2003; Li et al.2012; Bensby et al. 2014). The velocity dispersion for thin disk is smaller than 20 kms − (Bensby et al.2003, 2014). It is hard to simply compare our results to thin and thick disk, there are obviously differentvelocity dispersion for the north and south parts of our GASS, but entirely the velocity dispersion ofnorth groups are much larger than south groups. This asymmetry structure could be related to thedisk ripple/wave structure as expected or detected by previous work.(Carlin et al. 2014; Xu et al. 2015;Li et al. 2017; Wang et al. 2018, 2020) Dynamical properties comparison with disk population
Figure 5 shows E vs L z distribution for GASS members(red stars), disk stars(yellow and purpledensity background) choose from our K-giants sample which | Z | < kpc, Sagittarius stream membersselected from Yang et al. (2019)(green circles). From the density background, we can clear see twocomponents, yellow region are attributed by thin disk stars, the purple region are attributed by thickdisk stars. As we can see the E vs L z distribution of our GASS members are totally different fromthick disk and Sagittarius stream. It is located in an extended narrow region of thin disk stars buthave higher E and less L z value than most thin disk population. THE CHEMICAL ABUNDANCE FEATURES OF GASS4.1.
The metallicity distribution
The [M/H] for all K-/M-giants members spans a large distribution from − . to . , with the meanmetallicity around -0.56 dex, and metallicity dispersion around 0.22 dex as shown in Figure 6. Fromthe upper panel, we see that the total K- and M-giants have similar [M/H] distributions. The lowerpanel shows that the north and south structures distributions separately, which also have similar [M/H]distributions. For both K- and M-giants, of members the [M/H] value larger than − . Earlierworks about Mon/or other Anticenter structures were suggested they could be the remnants of dwarfgalaxies which merged in with the outer disk, but for the most Milky Way satellites or dwarf galaxies,the mean stellar metallicites much smaller than -1.5 dex(McConnachie 2012; Simon 2019). Consideringthe metallicity and E vs L z distribution of our GASS are far from the dwarf galaxies’ distribution, Sowe infer it is unlikely the remnants of dwarf galaxies merged in the MW outer disk. This [M/H] distribution is similar to that found by Chou et al. (2010) with high-resolution spectra of 21 M-giantsstars. This confirms that the groups in the north and south likely belong to one larger group.4.2.
The alpha-abundances distribution
Figure 7 presents the α -element abundances [ α /M] for LAMOST K- and M-giants obtained by SLAM (Zhang et al.2019). To compare with the Galactic disk and halo stars, we choose disk stars from LAMOST K-giants with | Z | < α /M] vs. [M/H] space),and for halo stars we select | Z | > α -element abundances [ α /M] as the thin disk, but are more metal-poor than typicalthin disk stars. This result is consistent with the continuation of metal-rich thin disk stars into the outerhalo (Haywood et al. 2016). It is also consistent with the α -abundance derived in Hayes et al. (2018)for the TriAnd substructure, which suggests that this feature is an ”extension” of the trend seen inthe disk. We infer that our GASS members may be part of the outer disk, representing a transition population between thethin disk, thick disk, and halo. Figure 8 shows K-giant stars selected in the same Galactic distance range as GASS,separated into two samples with 15 < r GC <
25 kpc and a more nearby “outer disk” sample between 10 < r GC < < | Z | <
12 kpc, 2 < | Z | < − < Z < | Z | < < | Z | < The thin and thick disk stars can naturally separate in [M/H] vs [ α /M] figure as show in figure 7, we selected the relative pure thinand thick disk stars within 1 σ distribution for the density distribution for distinct thick and thin disk clump in figure 7, and check wherethese pure thin and thick disk stars distribution in E vs L z figure. Li et al. are still some thin disk stars, but we can also see the clump of the thick disk star sequence.
The literature Chenget al. (2012) has suggested that the scale length of the thick disk is quite short, not much more than2 kpc, whereas traditionally a 3 kpc scale length was assumed. Our result seem to agree with that,duiring the range < r GC < kpc, | Z | < kpc are not related to the thick disk at all. In the lower panel,showing a distance range from 15 kpc to 25 kpc, we can see that the thin disk stars ( | Z | < < | Z | < α /M] > α /M] space.We infer this sequence could be the transition sequence between the thin disk-thick disk-halo. Comparing these twodistance ranges, we can see a clear variation in [M/H]-[ α /M] space within 2 < | Z | < < R GC <
15 kpc range with 2 < | Z | < ∼
19 kpc. In Figure 3, Haywood et al. (2016) shows thereis an inner-disk composed of thick disk and metal-rich thin-disk stars. The α abundance of 12 TriAnd substructuremember stars derived in Hayes et al. (2018) shows a similar distribution as our GASS in [ α /M] vs. [M/H] space; wenote that Hayes et al. also claimed TriAnd is an “extension” of the disk. We also compare our GASS with MWTD, the eccentricites e of MWTD are smaller than 0.4, all ourGASS e < . ; the metallicity of MWTD is similar to our GASS, previous detected MWTD are allin the solar neighborhood, but our GASS much further away. Anyway, except the distance range, thechemical abundance, kinematic feature and eccentricity of our GASS are all very similar to MWTD. Based on the evidence we have presented, we infer that the outer-disk sequence represents a different evolution,where the outer Milky Way still has more cold gas in present day and then maybe lower star formation efficiency thanthe inner disk. Our GASS members are could be part of the outer-metal-poor-disk stars, and the outer-disk couldextend to 30 kpc. DISCUSSION AND CONCLUSIONSBy combining IoM and FoF algorithms, Xue et al. (in preparation) selected 589 GASS K- and M-giant stars fromLAMOST DR5 in the Galactic anticenter region based on their similar kinematic and chemical abundance features.These stars cover a large range in r gc , but have similar angular momentum and energy distributions, which couldrelated to the previously identified substructures Mon, TriAnd, and A13, which we have collectively named GalacticAnticenter Substructure (GASS).Based on this sample, we present the observations including kinematic and chemical parameters of these stars inTable 1 and the calculated orbit parameters in Table 2. All the members are published in on-line readable catalogs.The GASS covers a large area of the sky, centered around the Galactic anticenter region on both sides of the MWdisk, in a Galactic longitude range from from 80 ◦ to 230 ◦ , while the Galactic latitude goes from − ◦ to 40 ◦ . Thiscoverage is in good positional alignment with the “band-like” structures detected in the Pan-STARRS map, especiallythe high lighted features B,C and D in Figure 3 of Slater et al. (2014).The velocity vector directions of GASS in the X-Y and Y-Z planes indicate that the GASS consists of two circleorbits on both sides of the MW disk which span a large distances from 15 kpc to 30 kpc. We can see clear velocitygradient on the Y-Z plane as shown in Figure 3, with the V Z pointing toward and outward from the mid-planeat different distances in the southern hemisphere. We also compared our GASS stars to the disk and Sagittarius streamin E - L z space. GASS members have a similar L-E distribution to the thin disk distribution.We also present the metallicity distribution of GASS. The total [M/H] distributions for K- and M-giant GASSmembers are similar. We did not find significant [M/H] differences for the GASS south and north rings. By comparingthe α -abundance with the Galactic components, the trend of [ α /M] is neither the same as the traditional Galactic disknor halo populations. The GASS distribution in [ α /M] vs [M/H] space is consistent with the continuation of metal-richthin disk stars into the outer halo (Haywood et al. 2016). It is also consistent with the α -abundance derived in Hayeset al. (2018) for the TriAnd substructure, which suggests that this feature is an ”extension” of the trend seen in thedisk.Our analysis shows that Mon, TriAnd, and A13 (and possibly including other stars near the Galactic anticenter)have similar kinematic and chemical features. These stars may be just part of the outer-disk, with this outer diskextending out to at least r gc ∼
30 kpc. GASS may have formed in the outer disk where there may still be more cold gas
30 20 10 0 10 20 X gc (kpc)2010010203040 Y g c ( k p c )
10 kpc20 kpc30 kpc l = 0 l = 90 l = 180 l = 270 K-GiantsM-GiantsSun r gc (kpc)2520151050510152025 Z g c ( k p c ) b = 90 b = 0 b = 90 Figure 2.
Spatial distribution of the candidates in the X-Y plane (left panel) and R-Z plane (right panel), where r gc = (cid:112) x + y .The Galactic center is at (0,0,0) and the Sun is at (-8.3,0,0) kpc. Galactic longitude and latitude (dashed) and curves at constantGalactocentric radius (solid) are shown. in the present day and then maybe lower star formation efficiency than the inner disk. We also can infer that theGASS stars may formed after the thick disk was formed because the molecular cloud density decreasedin the outer disk where the SFR might be less efficient than the inner disk.
This work is supported by NSFC grant No. 11988101, 11873052, 11890694, 11835057, 11703019,11503066, U1731129and 11703038, by the National Key R&D Program of China under grant No. 2019YFA0405500, and by China WestNormal University grants 17C053, 17YC507 and 16EE018. JLC acknowledges support from HST grant HST-GO-15228.001-A and NSF grant AST-1816196. R.A.M. acknowledges support from the Chilean Centro de Excelencia enAstrofisica y Tecnologias Afines (CATA) BASAL AFB-170002, and FONDECYT/CONICYT grant
Gaia
Gaia
Gaia
Multilateral Agreement.REFERENCES
An, D., Beers, T. C., Johnson, J. A., et al. 2013, ApJ, 763,65. doi:10.1088/0004-637X/763/1/65Antoja, T., Helmi, A., Romero-G´omez, M., et al. 2018,Nature, 561, 360Arifyanto, M. I., Fuchs, B., Jahreiß, H., et al. 2005, A&A,433, 911. doi:10.1051/0004-6361:20035829Bailer-Jones, C. A. L., Rybizki, J., Fouesneau, M.,Mantelet, G., & Andrae, R. 2018, AJ, 156, 58 Beers, T. C. & Sommer-Larsen, J. 1995, ApJS, 96, 175.doi:10.1086/192117Beers, T. C., Carollo, D., Ivezi´c, ˇZ., et al. 2012, ApJ, 746,34. doi:10.1088/0004-637X/746/1/34Beers, T. C., Chiba, M., Yoshii, Y., et al. 2000, AJ, 119,2866. doi:10.1086/301410Beers, T. C., Norris, & Sommer-Larsen,J. 1995, ApJS, 96,175 Li et al.
15 10 5 0 5 10 15 20 25 Y gc (kpc)15105051015 Z g c ( k p c )
15 10 5 0 5 10 15 20 25 Y gc (kpc)15105051015 Z g c ( k p c ) K-, M-giants -100-80-60-40-20020406080100 V Z ( k m s )
15 10 5 0 5 10 15 20 25 Y gc (kpc)15105051015 Z g c ( k p c ) K-giants -100-80-60-40-20020406080100 V Z ( k m s )
15 10 5 0 5 10 15 20 25 Y gc (kpc)15105051015 Z g c ( k p c ) M-giants -80-60-40-200204060 V Z ( k m s ) Figure 3.
Spatial distribution of the candidates in the Y-Z plane. In the upper left panel, the arrows represent the star ´ sinstantaneous direction of motion and the velocity amplitudes. In the other panels the color shows the mean V Z in each bin inY-Z space. We can clearly see V Z gradient along the Y axis. Considering that there could be systematic distance errors betweenK- and M-giants which might affect the oscillation, we illustrate the K- and M-giants mean V Z density maps separately in thelower panels, where the effect is still present.Beers, T. C., Norris, J. E., Placco, V. M., et al. 2014, ApJ,794, 58Bensby, T., Feltzing, S., & Lundstr¨om, I. 2003, A&A, 410,527. doi:10.1051/0004-6361:20031213Bensby, T., Feltzing, S., & Oey, M. S. 2014, A&A, 562,A71. doi:10.1051/0004-6361/201322631Bidelman, W. P. & MacConnell, D. J. 1973, AJ, 78, 687.doi:10.1086/111475Binney, J., & Sch¨onrich, R. 2018, MNRAS, 481, 1501Bird, S. A., Xue, X.-X., Liu, C., et al. 2019, AJ, 157, 104Bland-Hawthorn, J., Sharma, S., Tepper-Garcia, T., et al.2019, MNRAS, 486, 1167Carlin, J. L., DeLaunay, J., Newberg, H. J., et al. 2014,American Astronomical Society Meeting Abstracts
100 50 0 50 100 V los (km s ) N n o r m North KG ( = 20.33 km s , =44.08 km s )North MG ( = 26.87 km s , =32.32 km s )
100 50 0 50 100 V los (km s ) N n o r m South KG ( =17.91 km s , =17.31 km s )South MG ( =16.27 km s , =28.86 km s ) Figure 4.
Line-of-sight velocity distribution for K- and M-giants members of GASS in north and south hemisphere separately.The red and blue dash lines show the Gaussian distribution for each groups, the mean velocities and velocity dispersion showin the figure.Gilmore, G., Wyse, R. F. G., & Norris, J. E. 2002, ApJL,574, L39. doi:10.1086/342363G´omez, F. A., White, S. D. M., Marinacci, F., et al. 2016,MNRAS, 456, 2779Hammersley, P. L., & L´opez-Corredoira, M. 2011, A&A,527, A6Hayes, C. R., Majewski, S. R., Hasselquist, S., et al. 2018,ApJL, 859, L8Haywood, M., Lehnert, M. D., Di Matteo, P., et al. 2016,A&A, 589, A66Hernquist, L. 1990, ApJ, 356, 359Ibata, R. A., Irwin, M. J., Lewis, G. F., Ferguson,A. M. N., & Tanvir, N. 2003, MNRAS, 340, L21Ji, W., Cui, W., Liu, C., et al. 2016, ApJS, 226, 1Laporte, C. F. P., G´omez, F. A., Besla, G., et al. 2018,MNRAS, 473, 1218Laporte, C. F. P., Minchev, I., Johnston, K. V., & G´omez,F. A. 2019, MNRAS, 485, 3134Li, J., Newberg, H. J., Carlin, J. L., et al. 2012, ApJ, 757,151Li, J., Smith, M. C., Zhong, J., et al. 2016, ApJ, 823, 59Li, T. S., Sheffield, A. A., Johnston, K. V., et al. 2017,ApJ, 844, 74Li, Y.-B., Luo, A.-L., Du, C.-D., et al. 2018, ApJS, 234, 31Li, C. & Zhao, G. 2017, ApJ, 850, 25Liu, C., Deng, L.-C., Carlin, J. L., et al. 2014, ApJ, 790, 110Liu, C., Wang, Y.-G., Shen, J., et al. 2017, ApJL, 835, L18 L´opez-Corredoira, M., Cabrera-Lavers, A., Garz´on, F., &Hammersley, P. L. 2002, A&A, 394, 883Luo, A.-L., Zhang, H.-T., Zhao, Y.-H., et al. 2012, Researchin Astronomy and Astrophysics, 12, 1243Majewski, S. R., Schiavon, R. P., Frinchaboy, P. M., et al.2017, AJ, 154, 94Martin, J. C. & Morrison, H. L. 1998, AJ, 116, 1724.doi:10.1086/300568McConnachie, A. W. 2012, AJ, 144, 4.doi:10.1088/0004-6256/144/1/4Meisner, A. M., Frebel, A., Juri´c, M., & Finkbeiner, D. P.2012, ApJ, 753, 116Moitinho, A., V´azquez, R. A., Carraro, G., et al. 2006,MNRAS, 368, L77Momany, Y., Zaggia, S., Gilmore, G., et al. 2006, A&A,451, 515Morrison, H. L., Flynn, C., & Freeman, K. C. 1990, AJ,100, 1191Morrison, H. L., Helmi, A., Sun, J., et al. 2009, ApJ, 694,130. doi:10.1088/0004-637X/694/1/130Navarro, J. F., Frenk, C. S., & White, S. D. M. 1996, ApJ,462, 563Newberg, H. J., & Carlin, J. L. 2016, Astrophysics andSpace Science Library, 420,Newberg, H. J., Yanny, B., Rockosi, C., et al. 2002, ApJ,569, 245 Li et al. L z (10 km/s kpc)1.81.61.41.21.00.80.60.40.2 E ( k m / s ) d i s k s t a r s Figure 5.
E-L Z density distribution for Galactic disc stars (coloured density, the data is from the K giants catalogue with | Z | < [M/H] (dex) N n o r m KG ( = 0.57 dex, =0.26 dex)MG ( = 0.55 dex, =0.18 dex) [M/H] (dex) N n o r m North ( = 0.54 dex, =0.22 dex)South ( = 0.57 dex, =0.22 dex)
Figure 6.
Metallicity distribution for our GASS members. The left panel shows a comparison between the M giants (red line)and the K giants (blue dash line). The right panel shows comparison between the south and north samples. [M/H] (dex) [ / M ] ( d e x ) GASS M-giantsGASS K-giants
Thin diskThick disk
Halo stars d i s k s t a r s Figure 7. [ α /M] versus [M/H] distribution of stars from GASS (grey stars), compared to the thick disk, thin disk, and halostars (blue density). Li et al. [M/H] (dex) [ / M ] ( d e x )
10 kpc < r gc < 15 kpc [M/H] (dex) [ / M ] ( d e x )
15 kpc < r gc < 25 kpc Figure 8. [ α /M] versus [M/H] distribution of the K-giants sample. The upper panel shows all stars selected between 10 15 kpc but where we have subtracted the group members identified in XXXue19, and then split into samples of differentdistance to the Galactic plane: 7 < Z < 12 kpc, 2 < Z < − < Z < − − < Z < − − < Z < < r GC < 25 kpc. Yang, C., Xue, X.-X., Li, J., et al. 2019, ApJ, 880, 65York, D. G., Adelman, J., Anderson, J. E., et al. 2000, AJ,120, 1579. doi:10.1086/301513Zhao, G., Zhao, Y.-H., Chu, Y.-Q., et al. 2012, RAA, 12,723 Zhang, B., Liu, C., Deng,Licai. 2019, arXiv:1908.08677Zhong, J., Li, J., Carlin, L. Jeffery., et al. 2019,ApJS,244,8 Li et al. T a b l e . P a r a m e t e r s o f G A SSS t a r s L A M O S T a G a i a b T y p e R . A . D ec l. d ∆ d h r v ∆ h r v p m r a ∆ p m r a p m d ec ∆ p m d ec [ M / H ] ∆ [ M / H ][ α / M ] ∆ [ α / M ] d e g d e g k p c k p c k m s − k m s − m a s y r − m a s y r − m a s y r − m a s y r − d e x d e x d e x d e x L A M O S T K G . . . . . . . . − . . − . . . . 05 7150992872282645040005120 L A M O S T K G . . . . − . . − . . − . . − . . . . 07 7161702872538315852901632 L A M O S T K G . . . . − . . − . . − . . − . . − . . 07 1107119115451783231530112 L A M O S T K G . . . . − . . . . − . . − . . . . 07 12021581888486850788253056 L A M O S T K G . . . . − . . − . . − . . − . . . . 07 12162191901404634945395456 L A M O S T K G . . . . − . . − . . − . . − . . . . 06 13150422867487021995351040 L A M O S T K G . . . . − . . − . . − . . − . . − . . 05 13151802867075525473014528 L A M O S T K G . . . . − . . − . . − . . − . . . . 08 1609069375247964054278912 L A M O S T K G . . . . − . . − . . − . . − . . . . 08 7704197311445949992518912 L A M O S T K G . . . . − . . − . . − . . − . . . . a U n i q u e i d e n t i fi e r i n L A M O S T . b S o l u t i o n i d e n t i fi e r i n G a i a . ( T h i s t a b l e i s a v a il a b l e i n i t s e n t i r e t y i n m a c h i n e - r e a d a b l e f o r m . ) T a b l e . O r b i t a l P a r a m e t e r s o f G A SSS t a r s L A M O S T e ∆ e a ∆ a l o r b ∆ l o r b b o r b ∆ b o r b l a p o ∆ l a p o E ∆ E L ∆ L k p c k p c d e g d e g d e g d e g d e g d e g k m s − k m s − k m s − k p c k m s − k p c . . . . . . . . . . − . . . . 19 71509919 . . . . . . . . . . − . . . . 74 71617016 . . . . . . . . . . − . . . . 95 110711921 . . . . . . . . . . − . . . . 48 120215820 . . . . . . . . . . − . . . . 22 121621916 . . . . . . . . . . − . . . . 88 131504215 . . . . . . . . . . − . . . . 39 131518023 . . . . . . . . . . − . . . . 50 160906917 . . . . . . . . . . − . . . . 92 770419717 . . . . . . . . . . − . . . . ( T h i s t a b l e i s a v a il a b l e i n i t s e n t i r e t y i n m a c h i n e - r e a d a b l e f o r m ..