High-α-Metal-Rich stars in the LAMOST-MRS survey and its connection with the galactic bulge
Haopeng Zhang, Yuqin Chen, Gang Zhao, Jingkun Zhao, Xilong Liang, Haining Li, Yaqian Wu, Ali Luo, Rui Wang
aa r X i v : . [ a s t r o - ph . GA ] F e b Research in Astronomy and Astrophysics manuscript no.(L A TEX: ms2020-0464.tex; printed on February 24, 2021; 3:07)
High- α -Metal-Rich Stars in the LAMOST-MRS Survey and ItsConnection with the Galactic Bulge Haopeng Zhang , , Yuqin Chen , , Gang Zhao , , Jingkun Zhao , Xilong Liang , Haining Li ,Yaqian Wu , Ali Luo and Rui Wang CAS Key Laboratory of Optical Astronomy, National Astronomical Observatories, Beijing 100101,China; [email protected] School of Astronomy and Space Science, University of Chinese Academy of Sciences, Beijing 100049,China
Received 20xx month day; accepted 20xx month day
Abstract
We report the detection of a large sample of high- α -metal-rich stars on the lowgiant branch with . < logg < . dex in the LAMOST-MRS survey. This special groupcorresponds to an intermediate-age population of − Gyr based on the [Fe/H]-[C/N] di-agram and age-[C/N] calibration. A comparison group is selected to have solar α ratio atsuper metallicity, which is young and has a narrow age range around 3 Gyr. Both groupshave thin-disk like kinematics but the former shows slightly large velocity dispersions. Thespecial group shows a larger extension in vertical distance toward 1.2 kpc, a second peakat smaller Galactic radius and a larger fraction of super metal rich stars with [Fe/H] > . than the comparison group. These properties strongly indicate its connection with the outerbar/bulge region at R = 3 − kpc. A tentative interpretation of this special group is thatits stars were formed in the X-shaped bar/bulge region, close to its corotation radius, whereradial migration is the most intense, and brings them to present locations at 9 kpc and beyond.Low eccentricities and slightly outward radial excursions of its stars are consistent with thisscenario. Its kinematics (cold) and chemistry ([ α /Fe] ∼ . ) further support the formation ofthe instability-driven X-shaped bar/bulge from the thin disk. Key words:
Stars: late type — stars: abundances — stars: fundamental parameters, Galaxy:disk — Galaxy: evolution — Galaxy:structure
Stars with super metallicity of [Fe/H] > in the solar neighborhood and beyond are interesting targets toinvestigate the role of radial migration in the Galactic disk. Based on large spectroscopic surveys, severalworks (Kordopatis et al. 2015; Anders et al. 2017; Chen et al. 2019) detected hundreds of super metallicity H.P. Zhang et al. stars in the solar neighborhood. It has been suggested that these stars originate from the inner Galaxy andhave moved to present locations by radial migration first introduced by Sellwood & Binney (2002). TheGalactic bulge has its main component with super metallicity (Ness et al. 2013) and thus becomes thenatural origin of these stars. This connection is also suggested by theoretical chemo-dynamical modelsof Minchev, Chiappini & Martig (2013) and other simulation works. In this respect, the observationalproperties of stars with super metallicity will provide important information on the Galactic bulge.The formation of the Galactic bulge is, however, complicate and raises hot debates in recent years. Asstated in the review paper by Barbuy (2018), the bulge is previously thought to be a collapsed and rapidly-formed spheroidal old component due to early mergers in the Galaxy (Zoccali et al. 2008), while it isrecently recognized to be a bar formed from dynamical instabilities in the disk (Di Matteo et al. 2016). Thecolour-magnitude diagrams (CMD) in the bulge region (Zoccali et al. 2008; Clarkson et al. 2008) wouldfavor for a old population of 10 Gyr and the narrow sequence at the turn off region of the CMD allows littleroom for the existence of young stars with age less than 8 Gyr. By adopting an evolving age-metallicityrelation for stars in the bulge, Haywood et al. (2016) reanalyzed the bulge CMD of Clarkson et al. (2008)by isochrones with a spread of ages, and found that all stars with [Fe/H] > are younger than 8 Gyr, whichwould make up 50% of all the stars in the bulge. With the release of the infrared spectroscopic data of theAPOGEE survey, more works (Sit & Ness 2020; Bovy et al. 2019) favor for a secular evolution of thebulge with a mixing population and thus young and old stars co-exist in the inner region of the Galaxy. Inparticular, Bovy et al. (2019) presented unprecedented detailed maps of the kinematics, elemental abun-dances, and ages from the bulge region to the outer disk. They found a long bar formed at 8 Gyr ago with asize of 5 kpc, and in the bulge region of R < kpc different metallicity and kinematics are shown betweenstars inside and outside the bar, indicating a complicate history of the bulge region.In this paper, we investigate metal rich stars in the LAMOST mid-resolution spectroscopic survey, whichprovides us an extraordinary database of radial velocities, stellar parameters and elemental abundances fora large number of disk stars. In particular, we aim to derive their ages based on age-sensitive abundances,which is not available in the study of super metallicity stars from LAMOST low resolution spectroscopicsurvey by Chen et al. (2019). As compared with stellar position and kinematics, age, metallicity and el-emental abundances of metal rich stars bear important signatures of the bar/bulge population taking intoaccount the fact that they are not altered by secular evolution, such as radial migration and blurring. Oldmetal rich stars could be picked out and become the most interesting targets to investigate radial migrationprocess of the Galactic disk and provide new constraints on the bar/bulge’s formation. In 2018, the LAMOST Galactic spectroscopic survey (Zhao et al. 2006, 2012; Cui et al. 2012; Deng et al.2012) finished its first five-year regular survey with the low resolution mode (R ∼ R ∼ ) survey (MRS) at two wavelength ranges of ˚ A < λ < ˚ A (Blue band) and ˚ A < λ < ˚ A (Red band). The LAMOST-MRS survey is designed for achievingseveral scientific goals, e.g., Galactic archaeology, stellar physics, star formation, Galactic nebulae, etc (Liuet al. 2020). A large fraction of target stars has multiple observations over three times, which provide good α mr stars in the LAMOST survey 3 resulting spectra in order to detect variation in radial velocity to exclude binary stars and to select high-quality spectra in deriving reliable abundances of the sample stars. The LAMOST-MRS DR7 has released2,426,237 spectra have signal-to-noise (S/N) higher than 10 for both blue and red bands. Recently, Wang etal. (2020) provided radial velocities, stellar parameters ( T eff , logg , [Fe/H]) and abundances of 12 elements(C, N, O, Mg, Al, Si, Ca, Ti, S, Ni, Cr, Cu) by using the SPCANet neutral network (hereafter LAMOST-MRS-SPCANet DR7).Based on this database, we select our sample stars in the following steps. First, stars on the low giantbranch (LGB) with . < logg < . are picked out because reliable ages can be estimated from [C/N]ratios for these stars as shown in Hasselquist et al. (2019). Then we limit stars with [Fe/H] > − . sincewe focus on the study of metal rich stars in the Galactic disk. With the above two criteria, the sample hasa temperature range of < T eff < K . Note that Wang et al. (2020) derived abundances forLAMOST-MRS DR7 sample based on stars in common with the APOGEE-Payne DR14 database by Tinget al. (2019). With the release of APOGEE DR16, we compare these abundances with this more reliabledataset for stars with . < logg < . and derive the abundance calibrations for Fe, Mg, Si, Ca, Ti, C, N,O elements based on stars in common by linear fits to the data, which are shown in red lines of Fig. 1. Withthese calibrations, we obtain abundances in the same scale with APOGEE DR16 dataset.For the selected sample of stars, we calculate stellar positions, spatial velocities and orbital parametersbased on radial velocities from the LAMOST-MRS-SPCANet data, distances and proper motions from Gaia
DR2 (Gaia Collaboration 2018). The distance of r est from Gaia
DR2 is adopted and we limit stars withrelative error in parallax to be less than 10%. Orbital parameters are calculated from spatial velocity withthe help of Galactic potential
MWPotential2014 in Galpy by Bovy (2015). Details on the calculations arepresented in Chen et al. (2019) for super metal rich stars in the LAMOST low resolution survey. The solarradius is of 8.34 kpc, the solar circular velocity of km/s (Reid 2014) and the solar motion of (11.1,12.24, 7.25) km/s (Sch¨onrich & Binney 2009).Fig. 2 shows the distributions of the sample stars in the R − Z plane. Here we pick out two specificregions for later comparisons with blue circles corresponding to stars with | Z | > kpc as a representativeof the thick disk population and green circles being stars at very low Galactic-plane inner solar circle( | Z | < . kpc and R < . kpc) in the LAMOST-MRS-SPCANet DR7 dataset. [Fe/H] VERSUS [Mg/Fe]
DIAGRAM
Fig. 3 shows the [Fe/H] versus [Mg/Fe] diagram and its contour mapping for our sample stars. Two se-quences, the thin disk on the low [Mg/Fe] branch and the thick disk on the high [Mg/Fe] branch, are wellseparated by the red line of [Mg/Fe] = − . [Fe/H] + 0 . , which is drawn by passing through the sparseregion between the two branches at − . < [Fe/H] < − . . Interestingly, a special feature with [Fe/H] > and [Mg/Fe] ∼ . is shown in the contour map (lower panel) and its stars are marked in red dots (upperpanel). We note that this special group is similar as the high- α -metal-rich (h α mr) stars detected in the solarneighborhood by Adibekyan et al. (2012). Meanwhile, it could be the same feature as the upturn trend of[Si/Fe] for the most metal rich stars in the APOGEE survey by Haywood et al. (2013). The feature in thepresent work is clearer than that of APOGEE data probably because we limit the sample stars within the H.P. Zhang et al.
Fig. 1: The abundance calibrations of [ X/H ] for Fe, Mg, Si, Ca, Ti, C, N, O elements between the LAMOSTMRS DR7 and APOGEE DR16 based on common stars with . < logg < . and [Fe/H] > − . . Redsolid lines are linear fits to the data and red dash lines show the one-to-one relations.Fig. 2: The R − Z plane of the selected sample. For later comparisons, stars at the low- | Z | inner solar circlewith R < kpc and | Z | < . kpc are marked with green circles, and stars from the thick disk with | Z | > kpc are marked with blue circles. α mr stars in the LAMOST survey 5 Fig. 3: The [Fe/H] versus [Mg/Fe] diagram and its contour map for our sample stars. The red line with[Mg/Fe] = − . [Fe/H] + 0 . is drawn to separate the high-[Mg/Fe] and low-[Mg/Fe] sequences. Stars from the low- | Z | inner disk at R < kpc and | Z | < . kpc (green circles) and the thick disk at < | Z | < kpc (circles) are overplotted. LGB stage of . < logg < . dex so that stars have internally consistent abundances to show distin-guished features. Based on 27,135 red clump stars in the APOGEE survey, this group is clearly picked outas a separated feature by Ratcliffe et al. (2020) (Group 3, their Fig. 11). This supports the reality of thisspecial group in the LAMOST-MRS survey.We over-plot stars from the two selected regions (blue and green circles) on the contour map in the[Fe/H] versus [Mg/Fe] diagram. It shows that this special group forms an extension of the thick disk, sinceboth of them locate above the red line. However, only one star from the selected thick disk (blue circles)is found in this region, while quite a few stars from the very low- | Z | inner solar circle (marked by greencircles) belong to this special group. We have 3293 stars in the special group, and we select a comparisongroup, which includes 1565 metal rich stars with [Fe/H] > . and [Mg/Fe] < − . [Fe/H] + 0 . (belowthe red line). In the following section, we investigate the similarities and differences of age, chemical andkinematical properties between the two groups. H.P. Zhang et al. [C/N] ratios
Based on masses and ages derived by Martig et al. (2016) and abundances from APOGEE DR14,Hasselquist et al. (2019) interpreted the [C/N] versus [Fe/H] distributions as the age versus [Fe/H] trends,and found that LGB stars with . < logg < . dex have the most reliable ages derived from [C/N] ratios.Fig. 4 shows the [Fe/H] versus [C/N] diagram for our sample stars with red dots corresponding to h α mrstars in the special group. Two red lines indicate the locations of the old population of t = 9 Gyr and theyoung population of t = 2 . Gyr, which are derived from open clusters in APOGEE DR14 by Hasselquistet al. (2019). The special group lies below the upper red line of t = 9 Gyr but far above the lower red oneof t = 2 . Gyr, indicating an intermediate-age population of < t < Gyr. The comparison group liesaround the lower red line with its main part of ∼ Gyr, slightly above the t = 2 . Gyr line.We estimate ages for our sample stars based on the age calibration of log ( Age [ yrs ]) = 10 .
54 +2 . [C/N] by Casali et al. (2019), which was derived from open clusters in the Gaia-ESO and APOGEEsurveys. This calibration is valid for stars with − . < [Fe/H] < . , and our sample stars ([Fe/H] > )fits this requirement. However, the comparison group shows a peak at 1 Gyr by using this age calibration,which indicates a systematic deviation in [C/N] ratio between Casali et al. (2019) and the present workfrom the dataset by Wang et al. (2020). Therefore, we shift the [C/N] term upward by 0.15 dex in theage calibration in order to fit the peak of 2-3 Gyr of the age distribution for the comparison group. Thatis, we adopt a revised age calibration of log ( Age [ yrs ]) = 10 .
54 + 2 . [C/N] + 0 . . Note that thereis no deviation in [C/N] ratio between Hasselquist et al. (2019) and our work because both are based onAPOGEE DR14 dataset. Wang et al. (2020) derived abundances for LAMOST-MRS DR7 sample based onstars in common with the APOGEE-Payne DR14 database by Ting et al. (2019), and thus it is also based onAPOGEE DR14 dataset. This consistency seems to persist when we correct abundances to the same systemof APOGEE DR16 via calibrations in Fig. 1.Based on the revised age calibration, the special group shows a wide age range of − Gyr with apeak at − Gyr. Since the age calibration of Casali et al. (2019) does not take into account the metallicitydependence, which is clearly seen in open clusters of Hasselquist et al. (2019), we expect that the old end ofthe age distribution is overestimated at the solar metallicity where the highest [C/N] is found in the specialgroup. In view of this, we suggest that this special group has a upper age limit of 9 Gyr, rather than the − Gyr from the revised age calibration.The significant deviation of age by ∼ − Gyr between the special and the comparison groups isconsistent with the result of Sit & Ness (2020) that stars in the bulge are about twice as old (8 Gyr) as thosein the solar neighborhood (4 Gyr) for low- α stars, corresponding to a 4-Gyr-old difference. The age peakingat 7 Gyr of this special group is exactly the same as ∼ . Gyr of Bovy et al. (2019) (their Fig. 7) for thebar region based on APOGEE DR16, and is consistent with an age of ± Gyr by Sit & Ness (2020) forbulge stars within
R < . kpc in APOGEE DR14. Thus, the age distribution of the special group indicatesits connection with the bar/bulge, rather than the local-born young disk in the solar neighborhood. α mr stars in the LAMOST survey 7 Fig. 4: Upper: The [Fe/H] versus [C/N] diagram with the two red lines from Hasselquist et al. (2019)corresponding the ages of 2.5 Gyr and 9 Gyr based on open clusters. Lower: The age histograms of low(black line) and high [Mg/Fe] (red line) metal rich stars at [Fe/H] > based on the revised calibration of log ( Age [ yrs ]) = 10 .
54 + 2 . [C/N] + 0 . (from the original one by Casali et al. (2019)). Fig. 5 shows histograms of abundance ratios for six elements (Mg, C, O, Al, Si, Ti) between the special andthe comparison groups. Besides [Mg/Fe], other α -related elements, [O/Fe] and [Ti/Fe], as well as [Al/Fe] inthe former are overabundant. Since these ratios are all age sensitive (Delgado Mena et al. 2019), it furtherconfirms the age separation between the two groups. Meanwhile, the special group has wider abundancedistributions than the comparison group for all age-sensitive elements, indicating a larger age range thanthe comparison group. Note that [Si/Fe] distribution of the special group overlaps significantly with thecomparison group probably due to the uncertainty of Si abundances in both LAMOST-MRS-SPCANetDR7 and APOGEE DR14 dataset. Actually, Zasowski et al. (2019) revealed the presence of abundancetrends with temperature for Si and Ca elements. We do not include [Ca/Fe] in the comparison not onlybecause it has large uncertainties but also because it is not age-sensitive (Delgado Mena et al. 2019).The deviations in peak of these histograms between the two groups are 0.10 dex for [Mg/Fe], 0.16 dexfor [C/Fe] and 0.12 dex for [O/Fe]. The overabundant [Mg/Fe] of 0.10 dex is similar as that of the metalrich component in the bulge region with R < . kpc by Rojas-Arriagada et al. (2019). The overabundant[O/Fe] of . dex with respect to the comparison group is close to the average value of bar/bulge by Bovyet al. (2019) (their Fig. 5.). H.P. Zhang et al.
Fig. 5: The comparison of histograms of abundance ratios for Mg,C,O,Al,Si,Ti elements between the specialand the comparison groups. R , | Z | and [Fe/H] histograms The R, | Z | and [Fe/H] histograms between the special and the comparison groups are presented in Fig. 6.It shows that both groups locate around 9 kpc with the former extending towards lower values until 7 kpc.The comparison group has | Z | < . kpc and [Fe/H] < . in consistent with the origin of the young thindisk population, while the special group has main contributions from stars with | Z | > . kpc and withmore extending metallicity towards [Fe/H] ∼ . dex.If the comparison group represents the local-born thin disk population as indicated by young age and low | Z | < . kpc, then the local ISM in the solar neighborhood is enriched to [Fe/H] ∼ . at 3 Gyr ago. Thelocal ISM metallicity should be significantly lower at 7 Gyr ago when stars of the special group were born.The fact that this intermediate-old special group has even higher metallicity than the young comparisongroup highly suggests that it does not belong to the local region but comes from a high-SFR region, e.g. thebulge. According to Rojas-Arriagada et al. (2019), the bulge region does show a high [Mg/Fe] sequenceextending to super solar metallicity. Interestingly, they found that this metal rich component has a upperlimit of | Z | ∼ . kpc, similar as the | Z | distribution of the special group (until 1.2 kpc) in our sample.In addition, the X-shaped structure of the bulge has vertically extended feature to | Z | ∼ . kpc (Barbuy2018). It seems that this special group with similar | Z | extension is related with the X-shaped structure ofthe bulge region. Fig. 7 presents the R versus V φ diagrams for the comparison and the special groups as well as the thick diskrepresented by stars with | Z | > kpc. Two dense clumps are clear for the former and they become slightlysparse for the latter, while there is almost no any clump for stars with | Z | > kpc. The velocity of the specialgroup is more like the thin disk as most stars distributing above the red dash line of V φ ∼ km/s . Stars α mr stars in the LAMOST survey 9 Fig. 6: The R , | Z | and [Fe/H] histograms of the the special and the comparison groups.from the selected thick disk at | Z | > kpc show a wide distribution and more stars have V φ < km/s than stars with V φ > km/s . The V R and W LSR velocities have the same features as V φ . The low V φ dispersion of this special group indicates its kinematic colder than the group of high | Z | thick disk.This is consistent with the metal-rich component (mean [Fe/H] = 0 . ) of the bulge in Ness et al. (2013)who found this component is kinematically cold and has the thin-disk characteristics, typical for the outerbar/bulge as shown in Bovy et al. (2019).The distributions of eccentricity, R − R g and R apo − R peri are similar for both groups as presented inFig. 8. There is a low eccentricity distribution of e < . , | R − Rg | < kpc but quite a broad distributionof radial excursion distance of R apo − R peri (from 1 to 5 kpc). Again, a small fraction of stars of the specialgroup extend to higher values in these histograms, which indicates a outward displacement due to secularevolution. In short, the special group is kinematically cold, similar as the thin disk population, and has thesame | Z | distribution as the metal-rich component of the outer bulge and the X-shaped bar. Based on the above properties (age, Galactic location, kinematics and chemistry), we propose a tentativeinterpretation on the origin of the special group by investigating how it connects with the bulge, where(inner or outer region of the bulge) it comes from, and how its stars stray from their birth sites to presentlocations.In the past few years, some observational and theoretical works have discussed the connection betweenthese special stars and the inner disk. Using Gaia ESO survey, Kordopatis et al. (2015) pointed out thatthe metallicity of high- α stars in the solar neighbourhood probably up to . dex, and discussed that thesestars may be formed in the inner most region of the Milky Way. However, due to the lack of reliable agesof these stars, they could not draw reliable conclusions. Minchev, Chiappini & Martig (2013) showed thatmost of the stars with super solar metallicity were born in the inner disk according to the chemodynamicalsimulation, and almost all the stars with metallicity of . − . dex in the solar neighbourhood came from − kpc. Minchev et al. (2018) estimated the birth radius of field stars by a semi-empirical approach, andcalculated that the birth radius of high- α -metal-rich stars in their samples taken by the HARPS instrumentis basically less than kpc (see their Fig. 9). Fig. 7: The R versus V φ diagrams and the contour maps for the comparison (left), the special (middle)groups and the thick disk stars with | Z | > kpc (right).Fig. 8: The distributions of eccentricity, R − R g and R apo − R peri for the special and the comparison groups.According to Fragkoudi et al. (2018), the bulge is slightly metal-poor with [Fe/H] = − . dex (seetheir Fig. 13) in the innermost regions ( R < kpc), while the most metal rich component has the meanmetallicity of [Fe/H] = +0 . dex in the outer bulge where the bar is strong. However, Wegg et al. (2019)suggested that stars in the central bar (also the central bulge) are more metal-rich than their surrounding re-gion and thus suggested an opposite metallicity structure of the bulge. Based on APOGEE and Gaia surveys,Bovy et al. (2019) presented unprecedented detailed maps of the kinematics, elemental abundances, and α mr stars in the LAMOST survey 11 age of the bulge region, and found that the central bar tend to be slightly metal poor than the bulge regionoutside the bar, which is consistent with the result of Fragkoudi et al. (2018). In particular, they suggestedthat the highest metallicity region is in the outermost region of the bar/bulge, i.e. − kpc, where the agehas a peak of ∼ ∼ . dex (with respect to solar radius) as shown intheir Figure 5. All these properties are found for the special group in our sample, and thus we suggests thatit may originate from R = 3 − kpc, rather than the central bar/bulge ( R < kpc).The bar/bulge shows a X-shape (or boxy/peanut shape) structure in the X-Z diagram (Barbuy 2018),which has an extension to ∼ . kpc in | Z | direction. According to Di Matteo et al. (2016), the X-shapestructure is found to be more prominent in the metal-rich population than other components. In other words,stars in the most metal rich component of the bulge have a large extension in vertical distance, reaching 1.2kpc at the outer region of − kpc. The similar | Z | extension to 1.2 kpc for the special group in ourwork supports the suggestion that its stars come from the outer bar/bulge region at R = 3 − kpc and < | Z | < . kpc.Finally, it is interesting to investigate how stars in this special group move from their birth sites of − kpc to present locations peaking at − kpc (see Fig. 6). The low eccentricity (peaking at . − . ),small | R − R g | ( < kpc) and kinematical cold (small velocity dispersion) all indicate that it is unlikelythat blurring alone could cause a distance displacement of − kpc in radial direction without inducingsignificant velocity dispersions. Churning (i.e. radial migration) is, however, a promising mechanism tobring these stars to present locations in the sense that their birth sites of − kpc are close to the corotationradius (CR) of the Galactic bar, which locates at 4.7 kpc (Minchev, Chiappini & Martig 2013) or 5.5 kpc(Bovy et al. 2019), where radial migration is most intense. According to Minchev, Chiappini & Martig(2013), the incorporating spiral arms and bar structure could provide a very powerful stellar radial migration,leading to the migration distance as large as 5 kpc. Moreover, radial excursions between peri-center and apo-center distances ( R apo − R peri ) are of − kpc (Fig. 8), which allows for a radial displacement of these starsoutwards to larger distances. The low eccentricity of the special group is consistent with the scenario thatthese stars stray from the outer bar/bulge at R ∼ − kpc and moved outwards by radial migration, whichalters the guiding radius without changing the eccentricity. If this is the case, its thin-disk characteristics,i.e. kinematic-cold, intermediate-age and mildly-enhanced [Mg/Fe] ratio, would provide a support for thebar/bulge formation from the thin disk (Di Matteo et al. 2016). Based on LAMOST-MRS-SPCANet DR7 data, we selected a sample of stars with
S/N > and stellarparameters in the ranges of < T eff < K , . < logg < . and − . < [Fe/H] < . dex.In the [Fe/H] versus [Mg/Fe] diagram, two distinct groups are detected at super solar metallicity, a specialgroup with enhanced [Mg/Fe] of 0.1 dex and a comparison group with solar [Mg/Fe] ratios.Stellar ages are estimated from [C/N] ratios, and this special group corresponds to an intermediate-age population of − Gyr, while the comparison group is a young population peaking at − Gyr.The special group shows wide distributions with significant extension towards small R , large | Z | (until 1.2kpc) and high [Fe/H] as compared with the comparison group. The age-sensitive elements, [O/Fe], [C/Fe], [Al/Fe] and [Ti/Fe] in the special group are also overabundant as well as the enhanced [Mg/Fe] and [C/N]ratios. Both groups are thin-disk like in the R versus V φ diagrams and are kinematic cold as indicated bylow velocity dispersions. We connect this special group with the outer bar/bulge region based on similarproperties, i.e. the extending distributions of high | Z | and high [Fe/H], slightly overabundant [ α /Fe] ratios,thin-disk like kinematics and low velocity dispersions.By combining the above properties, we propose a tentative interpretation on the origin of the specialgroup. It was born at the outer bar/bulge of R ∼ − kpc, where is the most metal-rich componentwith an extending | Z | distribution to 1.2 kpc, typical for the X-shape structure at the outer bar/bulge. Radialmigration induced by the coupling between the bar and the spiral arms (Minchev, Chiappini & Martig 2013)could bring its stars from the birth sites of − kpc to present locations. If the bulge origin of this specialgroup is true, its kinematic-cold, intermediate-age and mildly-enhanced [Mg/Fe] would provide a supportfor the scenario that the bar/bulge is formed from the thin disk population as suggested by Di Matteo et al.(2016). The future 2-m Chinese Space Survey Telescope (CSST) project (Zhan et al. 2011) will provide agood opportunity to investigate the properties of the bulge and the mechanism of radial migration by findinga large amount of super solar metallicity stars in the inner disk. The preparation of the usage of data fromCSST project for this investigation is desirable, as already started for the study of the merging history byZhao & Chen (2021) via the chemical abundances of halo stars. Acknowledgements
References
Adibekyan, V. Z., Sousa, S. G., Santos, N. C., et al. 2012, A&A, 545, A32 3Anders, F., et al., 2017, A&A, 597, A30 1Barbuy B, Chiappini C. & Gerhard O, 2018, Ann. Rev. Astro. & Astroph. 56, 223 2, 8, 11Bovy, J. 2015, ApJS, 216, 29 3Bovy, J., Leung, H. W., Hunt, J. A. S., et al. 2019, MNRAS, 490, 4740 2, 6, 7, 9, 10, 11Casali G., Magrini L., Tognelli E. et al., 2019, A&A, 629, A62 6, 7Chen Y Q, Zhao G, Zhao J K, et al. 2019, AJ, 158, 249 1, 2, 3 α mr stars in the LAMOST survey 13mr stars in the LAMOST survey 13