Chemodynamics of the Milky Way and disc formation history: insight from the RAVE and Gaia-ESO surveys

0.21 MB·

by  G. Kordopatis

AstronomischeNachrichten,13January2016
Chemodynamics of the Milky Way and disc formation history:
insight from the RAVE and Gaia-ESO surveys
Georges Kordopatis1(cid:63),(cid:63)(cid:63),RosemaryF.G. Wyse2,andJamesBinney3
1 Leibniz-Institutfu¨rAstrophysikPotsdam(AIP),AnderSternwarte16,14482Potsdam,Germany
2 JohnsHopkinsUniversity,DeptofPhysicsandAstronomy,3400NCharlesStreet,Baltimore,MD21218,USA
3 RudolfPeierlsCentreforTheoreticalPhysics,KebleRoad,OxfordOX13NP,UnitedKingdom
6
1 ReceivedXXXX,acceptedXXXX
0 PublishedonlineXXXX
2

n Keywords Galaxy:abundances–Galaxy:disc–Galaxy:stellarcontent–Galaxy:stellarcontent–Galaxy:dynamics
a andkinematics
J
Multi-objectspectrographshaveopenedanewwindowontheanalysesofthechemo-dynamicalpropertiesofoldMilky
1
1 Waystars.Theseanalysesallowustotracebacktheinternalmechanismsandtheexternalfactorsthathaveinfluencedthe
evolutionofourGalaxy,andthereforeunderstandfundamentalaspectsofgalaxyevolutioningeneral.Here,wepresent

] recentresultsfromtheRAdialVelocityExperiment(RAVE)andtheGaia-ESOsurvey.ThesesurveysexploretheMilky
A Waypropertiesindifferentways,intermsofsamplesizeandselection,magnituderange,andspectralresolution.Wefocus
G hereon(i)thefirstdirectdetectionofevidenceforradialmigrationwithinthethindisc,providinginsightintothehistory
ofspiralstructureoftheMilkyWay,and(ii)thechemo-dynamicalcharacterisationofthemetal-weakthickandthindiscs,
.
h forwhichchemo-dynamicalmodelsstillhavedifficultiesinreproducing.
p
-
Copyrightlinewillbeprovidedbythepublisher
o
r
t
s 1 Introduction ever, large statistical data sets are needed. For more than a
a
[ decade,multi-objectspectroscopicsurveyshavebeengath-

Thedisentanglementoftheeffectsofpastaccretionevents ering data with continually improving spectral resolution
1
v and of secular mechanisms in the evolution of the Milky and/or for larger number of stars. Such surveys include
3 Way is a complex task (e.g. Rix & Bovy, 2013; Sellwood, the Geneva Copenhagen Survey (Nordstro¨m et al., 2004),
7 2014). Large samples of stellar velocities and metallicities the RAdial Velocity Experiment (RAVE, Steinmetz et al.,
6
are very valuable in this endeavour, containing the signa- 2006),theSloanExtensionforGalacticUnderstandingand
2
tures of past events (e.g. Eggen et al., 1962; Freeman & Exploration(SEGUE,Yannyetal.,2009),theAPOGalactic
0
. Bland-Hawthorn, 2002), with the caveat that the distribu- Evolution Experiment (APOGEE, Majewski et al., 2015),
1 tions functions (DF) of the different Galactic populations the Gaia-ESO Survey (GES, Gilmore et al., 2012) and the
0
often overlap and change as a function of location within LAMOSTexperiment(Zhaoetal.,2012).TheongoingGaia
6
1 theGalaxy.Inprincipleageistheparameterwiththemost mission(Perrymanetal.,2001)willprovideastrometricin-
: power to discriminate among models of our Galaxy’s his- formation for more than a billion stars with, in particular,
v
tory(e.g.Minchevetal.,2014).However,stellarageshave measurementsoftheparallaxesandpropermotions,inaddi-
i
X only recently started being derived with reasonable error tiontophotometricinformationandspectraforasignificant
r bars, and for only a relatively small sample of stars, and fractionoftheobservedtargets.
a
potential biases in the derived ages may still exist (Miglio
et al., 2013). The large-scale investigation of the chemo-
dynamical properties of the stars constituting the different
Galacticpopulationsisthereforestillthemostrobustwayto
understand how our Galaxy was formed and evolved. This
Here we review some recent results, obtained with
isachievedbydetectingandidentifyingthesignaturespre-
dicted by models that incorporate different underlying for- RAVE data release 4 (DR4) (Kordopatis et al., 2013a) and
GES DR2, concerning radial migration and the accretion
mationmechanisms.
history of the Galactic discs. These results are based on
Thesesignaturesareencodedinthecorrelationsamong
analysesofthepropertiesofthetailsofthemetallicitydis-
thespatial,chemical,andkinematicparametersofeachstel-
tributions of the thin and thick disc, i.e. either the super
lar population. In order to detect these signatures, how-
metal-rich (Sect. 2) or the metal-poor ends (Sect. 3). Such
(cid:63) OnbehalfoftheRAVEandGaia-ESOcollaborations. analyses of course require large samples. Our perspectives
(cid:63)(cid:63) Correspondingauthor:[email protected] andconclusionsarepresentedinSect.4.
Copyrightlinewillbeprovidedbythepublisher
2 G Kordopatis:DiscchemodynamicswithRAVEandGaia-ESO
Fig.1 Normalised histograms of the ratio between the observed radius and the mean orbital radius of the metal-rich
RAVE stars for different metallicity bins (columns) and different Galactocentric radius rings (top:[6.5,7.5]kpc, middle:
[7.5,8.5]kpc,bottom:[8.5,9.5]kpc).Thenumbersintheupperrightcornersofeachpanelindicatethesizeofthesample
usedtocomputethathistogram.
2 Propertiesofthemetal-richendofthe rich stars in RAVE’s 4th data release, and found that more
thindisc than half of those with [M/H] above 0.3 were in circular
orbits, i.e. with a derived mean orbital (guiding centre) ra-
dius close to the radial coordinate at which the star is ob-
Starsinthelocalthindiscthathavemetallicitiesofatleast
served (see Fig. 1). This is strong evidence that these stars
twice the Solar value (denoted as super metal-rich stars,
havereachedtheSolarneighbourhoodthroughsomemech-
SMR) carry important information concerning the secu-
anismthatchangedtheirorbitalangularmomentumbutdid
lar evolution of the disc. Indeed, it is highly unlikely that
not increase their orbital eccentricities, the most likely be-
these stars could have formed in the Solar neighbourhood,
ingradialmigrationinducedbytorquesoperatingattheco-
since the local inter-stellar medium (ISM) has only now
rotationresonanceofatransientspiralperturbation(seefor
reached Solar metallicities (Cartledge et al., 2006). These
example,Sellwood&Binney,2002).
non-locally born stars could not have been accreted either,
sinceasatellitegalaxycontainingstarsofsuchhighmetal- ThelocalSMRstarsareestimatedtohavetypicalages
licity is expected to have been of high mass (Kirby et al., of ∼ 6 Gyr, similar to the Sun. Assuming a fixed radial
2013) and therefore its accretion should have left signif- metallicitygradientequaltothatoftheISMtoday(e.g.Gen-
icant phase-space imprints that so far have not been de- ovali et al., 2014), the SMR stars should have been born
tected.TheSupermetal-richstarshavethereforebeenborn in the inner regions of the Galaxy, R ∼ 2kpc. The results
withinthediscandreachedtheirpresentpositioneither(i) showninFig.1thereforehighlightthatinteractionsbetween
throughepicyclemotion(thereforeonlypassingthroughthe stars and transient spiral patterns at their co-rotation reso-
Solar neighbourhood on a presumably outwards radial ex- nances have occurred over a significant part of the history
cursion around apocentre), or (ii) having radially migrated oftheMilkyWay,inordertobringastarbornintheinner
outwards,changingtheirmeanorbitalradiusfromtheinner disc out to the Solar neighbourhood (its present position).
Galaxytolargerradiithroughinteractionsatthecorotation Further, such metal-rich stars have been detected up to al-
resonanceoftransientspiral armsand/orthecentralbar of mostonekiloparsecabovetheGalacticplane.Thisputsfur-
the Galaxy (e.g. Minchev & Famaey, 2010; Scho¨nrich & ther constraints on the history of the spiral structure in the
Binney,2009;Sellwood,2014;Sellwood&Binney,2002). Galaxy,sincethegravitationalfieldfromthespiralsshould
The RAVE survey operated for ten years and gathered havebeenstrongenoughtoinfluencestarsattheseheights.
data for more than 4 × 105 old FGK stars. Other than The gravitational field of a spiral of radial wave number k
avoidingthelowestGalacticlatitudefields,RAVEobserved decreases exponentially with height above the plane, z, as
theSouthernskyhomogeneously.Kordopatisetal.(2015a) exp(−kz) (Binney & Tremaine, 2008), implying the need
analysedthespatialandorbitalpropertiesofthemostmetal- for spiral perturbation of small wave number (long radial
Copyrightlinewillbeprovidedbythepublisher
asnaheaderwillbeprovidedbythepublisher 3
wavelength) (cf. Solway et al., 2012). The requirement for the Milky Way experienced a relatively quiescent merger
stars to migrate over a significant fraction of the extent of historyduringthelast9−10Gyr(cf.Wyse,2001).
thethindiscsuggeststheseperturbationswerealsoofhigh
amplitude, meaning massive spiral arms (see Kordopatis
3.2 Themetal-weakthindisc
etal.,2015a,forfurtherdetails).
Usingthesecondinternaldata-releaseofthehigh-resolution
survey Gaia-ESO1, Kordopatis et al. (2015b) investigated,
3 Themetal-poorendsofthediscs:isthere
withanewapproach,themetallicityrangeswhereanover-
evidenceofpastaccretionevents?
lapoftwopopulations(namelythethinandthethickdisc)
could be detected. This was achieved by determining the
Thelowestmetallicitiesreachedbythethickandthindiscs amplitude of trends in the azimuthal velocities as a func-
arealsodirectlylinkedtothehistoryoftheMilkyWay.In- tion of the magnesium to iron ratio (as a tracer of α-
deed,thelowmetallicitytailofthethickdiscprovidescon- abundances), for narrow metallicity bins and for stars lo-
straints on the potential extragalactic origin of these stars, cated in the extended solar neighbourhood. The advantage
whereastheα-enhancementofthelow-metallicitytailofthe of such an analysis is that it is very sensitive to the true
thindisc-andhowitcompareswiththeα-enhancementof tailsofthemetallicitydistributionfunctionsoftheunderly-
the metal-rich tail of the thick disc - holds information on ingpopulationswithout,however,assuminganyfunctional
the possible gas accretion history of the Milky Way (e.g. shape for the MDF. Indeed, the value of the slope of the
Chiappinietal.,1997;Wyse,1995). correlatonbetweenazimuthalvelocityandα-enhancement,
∂V /∂[α/Fe],isexpectedtobenon-zeroformetallicitybins
φ
where more than a single population exist, provided that
3.1 Themetal-weakthickdisc
a given stellar (mono-abundance) population has a lag in
meanazimuthalvelocity(comparedtothelocalstandardof
Selecting those RAVE stars located between 1 and 2kpc
rest)whichissufficientlydifferentfromthatofthetheother
fromtheplaneandwithmetallicitiesbelow−1.5dex,Kor-
population(s)inthesamesample.
dopatisetal.(2013b)foundahomogeneouslyspatiallydis-
tributed rotating stellar population. The azimuthal velocity The results obtained for stars located in the extended
of this population was found to be correlated with metal- solar neighbourhood (7.5 < R < 8.5kpc and |Z| < 1kpc)
licity, with the amplitude of the dependence being mea- areshowninFig.2,wherethecorrelationsfortheotherve-
sured as ∂V /∂[M/H] ≈ 50kms−1dex−1. This equals the locity components are also plotted. From that figure, one
φ
correlation measured previously for stars in other surveys can see that the trends for VR and VZ are always, within
probing the canonical (more metal-rich) thick disc (Kor- theerrors,centredaroundzero.Thisindicatesthatthevari-
dopatis et al., 2013c, 2011; Lee et al., 2011), suggesting ousstellarpopulationspresentinthis(sub-)samplehavethe
that this low-metallicity population represents the metal- same mean VR and VZ (no expansion or contraction of the
weak thick disc (MWTD). The large statistical sample of populations). In contrast, the Vφ panel shows clear trends,
RAVEallowedthedetectionoftheMWTDdowntometal- indicativeofamixofpopulations.Inparticular,itisfound
licitiesaslowas−2dex.Finally,despitenothavingaccurate that the value of ∂Vφ/∂[α/Fe] is non-zero in the metallic-
enough α-abundances to draw conclusions about their ori- ityrange[−0.8,+0.2]dex.Furthermore,itismonotonically
gin, the MWTD stars in Kordopatis et al. (2013b) did not increasing for metallicities between [Fe/H] = −0.6 and
show any particular distinction compared to the halo stars [Fe/H]≈0.Thesefactshavebeeninterpretedasfollows:
ofthesample.Thetransitionbetweenthehaloandthethick
– The relative proportions of thin to thick disc stars
disc has recently been explicitly investigated by Hawkins changesasafunctionofmetallicity,reflectingthediffer-
etal.(2015)usingAPOGEEdata,findingnochemicaldif-
enceinthemeanmetallicitiesofeachpopulation.How-
ferences between the two populations. This confirms ear-
ever, there appears to be a constant fraction of thick to
lierresultsbasedonasmallersampleselectedfromRAVE, thin disc stars below [Fe/H] (cid:46) −0.5, where the trend
with follow-up high resolution spectroscopy (Ruchti et al., changes slope, and is consistently different than zero
2011).
downto−0.8dex.
These results highlight the smooth transition from the – Both the thin and the thick disc have highly non-
halo to thethick disc: the innerhalo, thought to havebeen Gaussian shapes for their metallicity distribution func-
formed in situ (e.g. Carollo et al., 2007; Cooper et al., tions. This result is in agreement with the existence of
2015)showsnochemicaldifferencescomparedtothelow- thesupermetal-richtailofthethindisc(Sect.2)andis
metallicitythickdisc,possiblysuggestingthereforethatthe compatible with the existence of the metal-weak thick
latter was also formed in situ. This would be in agreement disc(Sect.3.1).
withthedynamicalstudyofRuchtietal.(2015),whoinves- – Thethindiscreachesmetallicitiesaslowas−0.8dex.
tigated the distribution of orbital angular momenta of the – Thethickdiscreachessuper-solarmetallicities.
stars in the Gaia-ESO catalogue and concluded that there
was no evidence for accreted disc stars and therefore that 1 TheGaia-ESOinternalDR2comprisesroughly104stars
Copyrightlinewillbeprovidedbythepublisher
4 G Kordopatis:DiscchemodynamicswithRAVEandGaia-ESO
Fig.2 Correlationsbetweenthe[α/Fe]abundancesandthethreevelocitycomponentsfortheSolarvicinitystars(7.5<
R < 8.5kpc and |Z| < 1kpc) of the Gaia-ESO internal DR2. No trends are detected, within the errors, for V and V .
R Z
In contrast, a clear signal is measured for the azimuthal velocity, indicating that the thin disc extends down to at least
[Fe/H]≈−0.8dexandthethickdiscextendsuptoatleastSolarmetallicities.
As already noted at the beginning of Sect. 3, the α- theHipparcosvolumeandthusareverynearbyandsimilar
enhancementsofthemetal-weakthindiscandofthemetal- inmetallicitytotheSun.
rich thick disc give valuable insight into the history of gas Inthecomingyears,thesampleforwhichwewillhave
accretion into our Galaxy. For example, the analyses of astero-seismicagesandparallaxesandpropermotionmea-
Gilmore & Wyse (1986), Chiappini et al. (1997) or more surements from space observations will increase signifi-
recently Haywood et al. (2013), Snaith et al. (2014) and cantly. This will open new doors into our understanding
Bensby et al. (2014), suggest that the thick disc formed oftheformationoftheMilkyWay,suchasidentifyingthe
rapidly during an early, possibly turbulent, phase of the relicsofminormergersandallowingagestobedetermined
young Milky Way, and that at least part of the metal-poor for stars of different chemical compositions. This will in
thindiscformedsubsequently,outofeithernewlyaccreted turnallowustohavea“live”pictureofourGalaxyandun-
gas at the end of the star-forming epoch of the thick disc derstandthesequenceofeventsthathavetakenplaceduring
and/orfromgasexpelledfromthethickdisc. thelast10Gyr.
According to these scenarios, the results derived from
Acknowledgements. G.K.thankstheWilhelmundElseHeraeus-
theGaia-ESOdatasuggesttwopossibilities.Thefirstisan
Stiftung foundation for financial support in order to attend the
accretion of a significant mass of metal-poor gas, to lower “ReconstructingtheMilkyWay’sHistory:SpectroscopicSurveys,
the ISM’s metallicity before the start of the thin-disc star AsteroseismologyandChemodynamicalModels”meetingheldin
formation,inordertoreproducethelargemetallicityover- Bad Honnef in June 2015. We acknowledge the contributions of
lap of the discs (over 1dex). Alternatively, observations ourcolleaguesintheRAVEandGaia-ESOsurveys.
could also beexplained with discs formed inan inside-out
mode(likelycombinedwithanviolenteventpuffingupthe References
thickdisc),whereinpracticethemetal-richthickdiscwould
have been formed in a different place than the metal-poor Bensby,T.,Feltzing,S.,&Oey,M.S.2014,A&A,562,A71
thindisc,withoutanyneedofdecreasingthemetallicityof Binney,J.&Tremaine,S.2008,GalacticDynamics:SecondEdi-
theISMatanyradius. tion(PrincetonUniversityPress)
Carollo,D.,Beers,T.C.,Lee,Y.S.,etal.2007,Nature,450,1020
Cartledge,S.I.B.,Lauroesch,J.T.,Meyer,D.M.,&Sofia,U.J.
4 ConclusionsandPerspectives 2006,ApJ,641,327
Chiappini,C.,Matteucci,F.,&Gratton,R.1997,ApJ,477,765
Thedataalreadyavailableallowustogaininsightintothe Cooper, A. P., Parry, O. H., Lowing, B., Cole, S., & Frenk, C.
formationprocessesoftheMilkyWaydiscsandtheaccre- 2015,ArXive-prints
Eggen,O.J.,Lynden-Bell,D.,&Sandage,A.R.1962,ApJ,136,
tion history of our Galaxy. However, there are limitations,
748
such as the relatively large uncertainties in distances (ob-
Freeman,K.&Bland-Hawthorn,J.2002,ARA&A,40,487
tained through isochrone fitting) and proper motions (ob-
Genovali,K.,Lemasle,B.,Bono,G.,etal.2014,A&A,566,A37
tainedfromgroundbasedsurveys)leadingtouncertainties
Gilmore,G.,Randich,S.,Asplund,M.,etal.2012,TheMessen-
in3Dspacevelocitiesof40-80kms−1.Inaddition,individ-
ger,147,25
ual stellar ages are very poorly known (the relative errors Gilmore,G.&Wyse,R.F.G.1986,Nature,322,806
from isochrone fitting is 40 percent, if not more), and the Hawkins, K., Jofre´, P., Masseron, T., & Gilmore, G. 2015, MN-
starsforwhichwedohaveareliableageestimatearewithin RAS,453,758
Copyrightlinewillbeprovidedbythepublisher
asnaheaderwillbeprovidedbythepublisher 5
Haywood,M.,DiMatteo,P.,Lehnert,M.D.,Katz,D.,&Go´mez,
A.2013,A&A,560,A109
Kirby,E.N.,Cohen,J.G.,Guhathakurta,P.,etal.2013,ApJ,779,
102
Kordopatis, G., Binney, J., Gilmore, G., et al. 2015a, MNRAS,
447,3526
Kordopatis,G.,Gilmore,G.,Steinmetz,M.,etal.2013a,AJ,146,
134
Kordopatis,G.,Gilmore,G.,Wyse,R.F.G.,etal.2013b,MNRAS,
436,3231
Kordopatis,G.,Hill,V.,Irwin,M.,etal.2013c,A&A,555,A12
Kordopatis, G., Recio-Blanco, A., de Laverny, P., et al. 2011,
A&A,535,A107
Kordopatis,G.,Wyse,R.F.G.,Gilmore,G.,etal.2015b,A&A,
582,A122
Lee,Y.S.,Beers,T.C.,An,D.,etal.2011,ApJ,738,187
Majewski,S.R.,Schiavon,R.P.,Frinchaboy,P.M.,etal.2015,
ArXive-prints
Miglio, A., Chiappini, C., Morel, T., et al. 2013, MNRAS, 429,
423
Minchev,I.,Chiappini,C.,Martig,M.,etal.2014,ApJL,781,L20
Minchev,I.&Famaey,B.2010,ApJ,722,112
Nordstro¨m,B.,Mayor,M.,Andersen,J.,etal.2004,A&A,418,
989
Perryman,M.A.C.,deBoer,K.S.,Gilmore,G.,etal.2001,A&A,
369,339
Rix,H.-W.&Bovy,J.2013,A&ARev.,21,61
Ruchti,G.R.,Fulbright,J.P.,Wyse,R.F.G.,etal.2011,ApJ,737,
9
Ruchti,G.R.,Read,J.I.,Feltzing,S.,etal.2015,MNRAS,450,
2874
Scho¨nrich,R.&Binney,J.2009,MNRAS,399,1145
Sellwood,J.A.2014,ReviewsofModernPhysics,86,1
Sellwood,J.A.&Binney,J.J.2002,MNRAS,336,785
Snaith,O.,Haywood,M.,DiMatteo,P.,etal.2014,ArXive-prints
Solway,M.,Sellwood,J.A.,&Scho¨nrich,R.2012,MNRAS,422,
1363
Steinmetz,M.,Zwitter,T.,Siebert,A.,etal.2006,AJ,132,1645
Wyse,R.F.G.1995,inIAUSymposium,Vol.164,StellarPopu-
lations,ed.P.C.vanderKruit&G.Gilmore,133
Wyse,R.F.G.2001,inAstronomicalSocietyofthePacificCon-
ferenceSeries,Vol.230,GalaxyDisksandDiskGalaxies,ed.
J.G.Funes&E.M.Corsini,71–80
Yanny,B.,Rockosi,C.,Newberg,H.J.,etal.2009,AJ,137,4377
Zhao,G.,Zhao,Y.-H.,Chu,Y.-Q.,Jing,Y.-P.,&Deng,L.-C.2012,
ResearchinAstronomyandAstrophysics,12,723
Copyrightlinewillbeprovidedbythepublisher