The Arcturus Moving Group: Its Place in the Galaxy
Mary E. K. Williams, Ken C. Freeman, Amina Helmi, RAVE collaboration
aa r X i v : . [ a s t r o - ph ] O c t The Galaxy Disk in Cosmological ContextProceedings IAU Symposium No. 254, 2008J. Andersen, J. Bland-Hawthorn and B. Nordstr¨om, eds. c (cid:13) The Arcturus Moving Group:Its Place in the Galaxy
Mary E. K. Williams , , Ken C. Freeman , Amina Helmi and theRAVE collaboration Mt Stromlo Observatory, Cotter Road, Weston Creek, ACT 2611, Australia, Astrophysikalisches Institut Potsdam, An der Sternwarte 16, D-14482, Potsdam, Germany, Kapteyn Institute, P.O. Box 800, 9700 AV Groningen, the Netherlandsemail: [email protected] ;email: [email protected] ;email: [email protected]
Abstract.
The Arcturus moving group is a well-populated example of phase-space substructurewithin the disk of our Galaxy. With its large rotational lag ( V = −
100 kms − ), metal-poor nature([Fe/H] ∼ − .
6) and significant age (10 Gyr) it belongs to the Galaxy’s thick disk. Traditionallyregarded as the remains of a dissolved open cluster, it has recently been suggested to be aremnant of a satellite accreted by our Galaxy.We confirm via further kinematic studies using the Nordstr¨om et al. (2004), Schuster et al.(2006) and RAdial Velocity Experiment surveys (RAVE, Steinmetz et al. 2006) the existenceof the group, finding it to possibly favour negative U velocities and also possibly a solar-circlephenomenon. We undertook a high-resolution spectroscopic abundance study of Arcturus groupmembers and candidates to investigate the origin of the group. Examining abundance of Fe,Mg, Ca, Ti, Cr, Ni, Zn, Ce, Nd, Sm and Gd for 134 stars we found that the group is chemicallysimilar to disk stars and does not exhibit a clear chemical homogeneity.The origin of the group still remains unresolved: the chemical results are consistent witha dynamical origin but do not entirely rule out a merger one. Certainly, the Arcturus groupprovides a challenge to our understanding of the nature and origin of the Galaxy’s thick disk. Keywords.
Galaxy: abundances, galaxy: kinematics and dynamics, galaxy: structure
1. Introduction
The Arcturus moving group was discovered by Eggen who over the years gathered alist of stars kinematically associated with Arcturus (Eggen 1971, 1998 and referencestherein). Eggen based group membership mainly upon the common V space velocity, thecomponent of stellar motion relative to the Local Standard of Rest (LSR) in the directionof rotation. For the Arcturus moving group the stars lag the LSR with V = ∼ −
100 kms − .While Eggen’s analysis was somewhat controversial as he adjusted the stellar parallaxes toreturn a tight V -velocity relation, a posteriori justification was obtained by a tight colour-magnitude relation along an isochrone with age τ &
10 Gyr and metallicity [Fe/H] ∼ − .
2. Abundance study
Candidate Selection
In light of the possible extra-Galactic origin of the Arcturus group there was the possi-bility that members could have been missed by Eggen as they might not satisfy his strictselection criterion. So in addition to Eggen’s list of Arcturus stars we selected group can-didates from the RAVE (Steinmetz 2006), Nordstr¨om et al. (2004), Beers et al. (2000) andNorris (1986) studies. To develop selection criteria inclusive of both formation scenarioswe performed N -body simulations of the formation of the Arcturus group via dissolutionof a progenitor in a static Galactic potential with a range of progenitor masses. Thesimulations follow those presented in Helmi (2006) and Navarro (2004) where a satelliteor cluster with an orbit similar to that of the Arcturus is disrupted in the potential of theGalaxy. The largest of these, the dissolved dwarf spheroidal galaxy, was used to developthe “banana” criterion selecting stars that are near or at the apocentre of their orbitsfalling within a banana-shaped region in the U V plane around V = −
100 kms − . Also,we restricted our study to stars on disk-like orbits with | W | <
100 kms − and with smallvelocity errors. Some stars from the thick and thin disks were included for comparativepurposes. A total of 134 stars from 190 observations were analysed in our study.2.2. Methodology
Elemental abundances were derived for our candidate stars by performing a Local Ther-modynamic Equilibrium (LTE) analysis with the MOOG code (Sneden 1973) on ourhigh resolution, high signal-to-noise UCLES data obtained in three observing periods atthe AAT from August 2003 - November 2006. The first of these runs was obtained byGayandhi de Silva and kindly granted to us for analysis. To measure equivalent widths weused the new DAOSPEC program which automatically fits Gaussian profiles to lines ina spectrum (Pancino & Stetson 2008). However, it was necessary to alter the DAOSPECcode with regards to continuum fitting of the spectra; most of our data is in the blue andso very crowded and it was found that DAOSPEC generally set the continuum too low.The DAOSEC continuum fitting was therefore disabled and hand fits performed. The linelist was compiled from the literature utilising laboratory log gf values where available.Abundances were derived for Fe, Mg, Ca, Ti, Cr, Ni, Zn, Ce, Nd, Sm and Gd for our134 stars using primarily spectra in the blue. For a subset of these stars we also have reddata from which we obtained abundances for Na, Mg, Al and Si. Here we present onlythose elements or lines for which hyperfine and isotopic splitting need not be considered.Stellar parameters were calculated using ‘physical’ and ‘spectroscopic’ approaches.The former involved calculating T eff and log g from photometric and astrometric data, he Arcturus Moving Group T eff and log g respectively. In both cases the microturbulence is found by requiring that there is nodependence of abundance on line strength. In this short report we only include thespectroscopic results as they yielded better agreement with abundance for stars in com-mon with the studies of Reddy et al. (2003, 2006) and Bensby et al. (2003, 2005). Thedifference in abundance between stars in common with those studies and our own are < [Fe/H] this study − [Fe/H] Reddy > = 0 .
09 with a standard deviation of σ = 0 .
05, andfor < [Fe/H] this study − [Fe/H] Bensby > = 0 .
05 with a standard deviation of σ = 0 . Results
In Section 3 we will see that the Arcturus group is an over-density in a very narrowvelocity range around V = −
100 kms − , with a standard deviation of only σ V = 3 kms − .We therefore revert here to a selection mimicking Eggen’s initial criteria, selecting asArcturus candidate stars those that are ±
10 kms − from this mean V velocity. Also,to compare the group against the background thick disk stars we include stars with V < −
50 kms − . Figure 1 shows selected abundance results for the Arcturus candidatessuperimposed on the background stars. The results from the Reddy et al. (2003, 2006)and Bensby et al. (2003, 2005) studies are included to further emphasise the generaltrends of field stars and to increase the number of candidate stars. The largest of thesystematic differences between these studies and our results were accounted for by usingthe zero-point offsets derived from common stars to shift their results to our scale.We see clearly in these plots that within the limit of our abundance errors there isno distinguishing features in abundance between the Arcturus group stars and those ofthe surrounding disk. While Nd suggests a possible clustering of abundance, this is notcorroborated by the α - or other elements. Also, the apparently tight CMD relation ofEggen now seems like a sampling from a old population of similar age, i.e., the thickdisk, as both the Arcturus candidates and the background stars lie reasonably tightlyalong the same isochrone. Employing different selection criteria for the group, such as ourbanana selection criteria mentioned above or selecting Eggen’s original Arcturus groupcandidates yields the same result: we are unable to find clear distinguishing features inabundance for the Arcturus group when compared to background disk stars.These results beg the question: does the Arcturus moving group exist at all as an over-density in phase space ? So before drawing any further conclusions we turn to kinematicstudies to confirm the group’s existence and define it kinematically.
3. Kinematic study
While the Geneva-Copenhagen catalog provided a significant number of Arcturus can-didates, it is not the ideal source because it has relatively few stars at the high-velocity ofthe group. However, the recent study of solar-neighbourhood metal-poor stars by Schus-ter et al. (2006) provides a wealth of stars in the thick disk. This study was not availableto us at the time of choosing candidates for our abundance study, however we can nowuse it to investigate the phase-space structure in the vicinity of Arcturus. Note that thereare kinematical selection biases in the Schuster data set towards high-proper motion starswhich means that stars with a V velocity nearer to the sun ( V ∼
0) are under-represented.However, the region around the Arcturus group’s velocity is not affected by these biases.Figure 2 displays plots of kinematic and metallicity plots for the combined Nordstr¨omand Schuster data sets where stars in common appear only once. In this diagram we see Mary E. K. Williams, Ken C. Freeman, Amina Helmi and the RAVE collaboration −200 −150 −100 −50 0 50V (km/s)−200−1000100200 U ( k m / s ) −200 −100 0 100 200U (km/s)−150−100−50050100150 W ( k m / s ) M V −2.0 −1.5 −1.0 −0.5 0.0 0.5[Fe/H]−0.20.00.20.40.6 [ S i / F e ] −2.0 −1.5 −1.0 −0.5 0.0 0.5[Fe/H]−0.20.00.20.40.6 [ T i / F e ] −2.0 −1.5 −1.0 −0.5 0.0 0.5[Fe/H]−0.20.00.20.40.6 [ α / F e ] −2.0 −1.5 −1.0 −0.5 0.0 0.5[Fe/H]−0.4−0.20.00.20.4 [ N i / F e ] −2.0 −1.5 −1.0 −0.5 0.0 0.5[Fe/H]−1.0−0.50.00.51.0 [ N d II/ F e ] −2.0 −1.5 −1.0 −0.5 0.0 0.5[Fe/H]−1.0−0.50.00.51.0 [ S m II/ F e ] Figure 1.
The
V U , UW planes, CMD and abundance ratio plots for the Arcturus groupcandidates (blue circles and red diamonds) selected from stars with V < −
50 kms − . Stars fromthis study are designated by circles, while the amalgamated Reddy and samples are diamonds.The overlaid Padova isochrone in the CMD has 12 . Z = 0 . an over-density at V ∼ −
100 kms − extending from 0 < [Fe/H] < −
1. This feature isparticularly prominent at lower metallicities. Schuster et al. (2006) interpreted this over-density as the thick disk being split into two components, with the split indicative ofthe thick disk consisting of merger debris. They link it with the Gilmore, Wyse & Norris(2002) feature but do not associate it to the Arcturus group. We see however that the‘second component’ of the thick disk that they identify is clearly at the Arcturus groupvelocity. Combining their data with the Nordstr¨om sample shows that the over-densityis perhaps not as localised in [Fe/H] as their results indicated.We see also that there is a suggestion of a bias towards negative U , which is coincidentwith the favoured U of the Hercules group. From fits to the generalised histogram inthe V -velocity we find that the group is centred around µ V = −
102 kms − with a verynarrow velocity dispersion of σ V = 2 − − .We have also investigated the Arcturus group in RAVE data, utilising the internalrelease of October 2007 which contains 191,170 radial velocity measurements as wellas stellar parameters for a significant number of stars (see Zwitter et al. 2008). Thestellar parameters enabled us to select those stars associated with the helium-burning he Arcturus Moving Group −200 −100 0 100 200U (km/s)−200−150−100−50050 V ( k m / s ) Arcturus −200 −150 −100 −50 0 50V (km/s)−4−3−2−101 [ F e / H ] Arcturus
Figure 2.
The Arcturus group as seen in the combined Nordstr¨om et al. (2004) and Schusteret al. (2006) data set. red clump. The K -band magnitude of the clump, while being relatively unaffected byextinction, has also been shown observationally to be relatively independent of metallicityand age (Pietrzy´nski, Gieren & Udalski 2003, Alves 2000). Thus, combining RAVE datawith photometric and astrometric data we were able to derive reliable distances and U V W velocites for a sample of 16,000 red clump giants. In this data set we have foundthat the Arcturus group seems to be more pronounced in the solar circle, i.e., for thosewith a Galactocentric radius similar to the sun (for full details see Williams et al. 2008).If this is indeed the case, such behaviour could be possibly be explained by a dynamicalorigin of the group.
4. Discussion
Venn et al. (2004) demonstrated that current-day satellite dSph galaxies of the MilkyWay are α -poor when compared to the Galactic stars of similar [Fe/H], indicating acomparatively slow star formation history (Matteucci 2003). So the similarity betweenthe Arcturus group members and surrounding disk weighs against the accretion debrisorigin of the group if we expect the Arcturus progenitor to be similar to current day dSphs .However, from Tamura, Hirashita & Takeuchi (2001) we see that a satellite progenitor ofArcturus would have had to have a substantial mass (similar to the LMC) to be enrichedto [Fe/H]=-0.6. Therefore, it is possible that a such a large dwarf galaxy had a muchhigher rate of star formation than the remaining satellites of the Milky Way. Also, thesimulations of Abadi (2003) of the formation of a Milky Way-like galaxy in the ΛCDMcosmogony gave 60% of the thick disk as tidal debris. So our results showing the Arcturusgroup as being similar to disk stars does not entirely exclude the possibility that it, andthe surrounding thick disk, are tidal debris. However, it would seem remarkable thata disparate group of satellites could conspire to have such similar abundances; a clearpicture of the chemical evolution of these satellites to produce the observed abundancepatterns has yet to be drawn.If we instead suppose an in situ formation of the thick disk, such as that proposedby Brook et al. (2005) from gas-rich mergers, the chemical similarity of the Arcturusgroup to the surrounding disk favours in situ formation of the group as well. However,from our simulations we find that even the largest open clusters (7 × M ⊙ ) could notproduce a discernible over-density in the solar neighbourhood after 10 Gyr of evolution(see Williams et al. (2008)). Furthermore, stars in the Arcturus velocity range do not Mary E. K. Williams, Ken C. Freeman, Amina Helmi and the RAVE collaborationexhibit the chemical homogeneity expected from a dissolved star forming event (e.g. see deSilva et al. (2007)). Instead we see that our results are similar to Bensby et al. (2007) forthe Hercules moving group where they found that the group is chemically indistinct fromthe disk. This supports the dynamic origin of the thick/thin disk Hercules group, whichis thought to arise due to the Outer Lindblad Resonance with the Galactic bar (Dehnen2000, Fux 2001). We therefore wonder, could this also be the case for the Arcturus group?This scenario could be supported by the Arcturus group seemingly mimicking Herculesin being asymmetrical in U and exhibiting a dependence on Galactocentric radius. Also,the V velocity of Arcturus is coincident with that of the 6:1 OLR of the bar at the solarposition. However, how such a resonance could produce an over-density, affecting starsthat spend so much time out the Galaxy’s plane is not currently understood.We are continuing our investigations of the Arcturus group. In our upcoming paper wewill present our abundance results in full, as well as including those for extra elements.We will give also present the full kinematic results, comparing them to our simulations ofvarious progenitor scenarios and explore the resonance possibility further. The Arcturusgroup’s origin is an enigma still yet to be solved, but the clues are accumulating. References
Abadi, M.G., Navarro, J.F., Steinmetz, M. & Eke, V.R. 2003,
ApJ , 597, 21Alves, D.R. 2000,
ApJ , 539, 732Beers, T.C., Chiba, M., Yoshii, Y., Platais, I., Hanson, R.B., Fuchs, B. & Rossi, S. 2000, AJ ,119, 2866Bensby, T., Feltzing, S. & Lundstr¨om, I. 2003, A&A , 410, 527Bensby, T., Feltzing, S., Lundstr¨om, I. & Ilyin, I. 2005,
A&A , 433, 185Bensby, T., Oey, M.S., Feltzing, S. & Gustafsson, B. 2007,
ApJ , 655, L89Brook, C.B., Gibson, B.K., Martel, H. & Kawata, D. 2005,
ApJ , 630, 298De Silva, G.M., Sneden, C., Paulson, D.B., Asplund, M., Bland-Hawthorn, J., Bessell, M.S. &Freeman, K.C. 2006, AJ ,131, 455Dehnen, W. 2000, AJ , 119, 800Eggen, O. 1971, PASP , 83, 271Eggen, O. 1996, AJ , 112, 1595Eggen, O. 1998, AJ , 115, 2397Fux, R. 2001, A&A , 373, 511Gilmore, G., Wyse, R.F.G. & Norris, J.E. 2002,
ApJ , 574, L39Helmi, A., Navarro, J.F., Nordstr¨om, B., Holmberg, J., Abadi, M.G., and Steinmetz, M. 2006,
MNRAS , 365, 1309Matteucci, F. 2003,
Ap&SS , 284, 539Navarro, J.F., Helmi, A. & Freeman, K.C. 2004,
ApJ , 601, L43Nordstr¨om, B., Mayor, M., Andersen, J., Holmberg, J., Pont, F., Jørgensen, B.R., Olsen, E.H.,Udry, S. & Mowlavi, N. 2004,
A&AS , 418, 989Norris, J. 1986,
ApJS , 61, 667Pacino & Stetson 2008, in preparationPietrzy´nski, G., Gieren, W., and Udalski, A. 2003, AJ , 125, 2494.Reddy, B.E., Lambert, D.L. & Allende Prieto, C. 2006, MNRAS , 367, 1329Reddy, B.E., Tomkin, J., Lambert, D.L. & Allende Prieto, C. 2003,
MNRAS , 340, 304Schuster, W.J., Moitinho, A., M´arquez, A., Parrao, L. & Covarrubias, E. 2006,