1001.0026.txt

arXiv:1001.0026v1  [astro-ph.SR]  30 Dec 2009Detectionof solar-likeoscillations from Keplerphotometry ofthe open
cluster NGC 6819
DennisStello,1Sarbani Basu,2HansBruntt,3Benoˆ ıt Mosser,3Ian R. Stevens,4
TimothyM.Brown,5Jørgen Christensen-Dalsgaard,6Ronald L. Gilliland,7Hans Kjeldsen,6
Torben Arentoft,6J´ erˆ omeBallot,8CarolineBarban,3TimothyR. Bedding,1WilliamJ. Chaplin,4
YvonneP. Elsworth,4Rafael A.Garc´ ıa,9Marie-Jo Goupil,3SaskiaHekker,4Daniel Huber,1
SavitaMathur,10Søren Meibom,11Reza Samadi,3VinothiniSangaralingam,4
Charles S. Baldner,2KevinBelkacem,12KatiaBiazzo,13Karsten Brogaard,6
Juan Carlos Su´ arez,14Francesca D’Antona,15Pierre Demarque,2LisaEsch,2NingGai,2,16
Frank Grundahl,6YvelineLebreton,17Biwei Jiang,16NadaJevtic,18ChristofferKaroff,4
AndreaMiglio,12JoannaMolenda- ˙Zakowicz,19JosefinaMontalb´ an,12ArletteNoels,12
Teodoro RocaCort´ es,20,21Ian W. Roxburgh,22AldoM. Serenelli,23VictorSilvaAguirre,23
ChristiaanSterken,24Peter Stine,18Robert Szab´ o,25AchimWeiss,23WilliamJ. Borucki,26
DavidKoch,26JonM. Jenkins27– 2 –
1SydneyInstituteforAstronomy(SIfA),SchoolofPhysics,U niversityofSydney,NSW2006,Australia
2DepartmentofAstronomy,YaleUniversity,P.O.Box 208101, New Haven,CT 06520-8101
3LESIA,CNRS,Universit´ ePierreetMarieCurie,Universit´ eDenisDiderot,ObservatoiredeParis,92195Meudon,
France
4SchoolofPhysicsandAstronomy,UniversityofBirmingham, Edgbaston,BirminghamB152TT,UK
5LasCumbresObservatoryGlobalTelescope,Goleta,CA 93117 ,USA
6DepartmentofPhysicsandAstronomy,AarhusUniversity,80 00AarhusC,Denmark
7SpaceTelescopeScienceInstitute,3700San MartinDrive,B altimore,Maryland21218,USA
8Laboratoired’AstrophysiquedeToulouse-Tarbes,Univers it´ edeToulouse,CNRS,14avE.Belin,31400Toulouse,
France
9Laboratoire AIM, CEA/DSM-CNRS, Universit´ e Paris 7 Didero t, IRFU/SAp, Centre de Saclay, 91191, Gif-sur-
Yvette,France
10IndianInstituteofAstrophysics,Koramangala,Bangalore 560034,India
11Harvard-SmithsonianCenterforAstrophysics,60GardenSt reet,Cambridge,MA,02138,USA
12Institutd’AstrophysiqueetdeG´ eophysiquedel’Universi t´ edeLi` ege,17All´ eedu6Aoˆ ut,B-4000Li` ege,Belgium
13ArcetriAstrophysicalObservatory,LargoE.Fermi5,50125 ,Firenze,Italy
14InstitutodeAstrof´ ısicadeAndaluc´ ıa(CSIC),Dept. Stel larPhysics,C.P. 3004,Granada,Spain
15INAF -Osservatoriodi Roma,via diFrascati 33,I-00040,Mon teporzio,Italy
16DepartmentofAstronomy,BeijingNormalUniversity,Beiji ng100875,China
17GEPI,ObservatoiredeParis,CNRS, Universit´ eParisDider ot,5Place JulesJanssen,92195Meudon,France
18Departmentof Physics& EngineeringTechnology,Bloomsbur gUniversity,400East SecondSt, BloomsburgPA
17815,USA
19AstronomicalInstitute,UniversityofWrocław,ul.Kopern ika11,51-622Wrocław,Poland
20DepartmentodeAstrof´ ıca,Universidadde LaLaguna,38207 LaLaguna,Tenerife,Spain
21InstitutodeAstrof´ ıcadeCanarias,38205La Laguna,Tener ife,Spain
22QueenMaryUniversityofLondon,Mile EndRoad,LondonE14NS ,UK
23MaxPlanckInstituteforAstrophysics,KarlSchwarzschild Str. 1,GarchingbeiM¨ unchen,D-85741,Germany
24Vrije UniversiteitBrussel, Pleinlaan2,B-1050Brussels, Belgium
25KonkolyObservatory,H-1525Budapest,P.O. Box67,Hungary
26NASA AmesResearchCenter,MS 244-30,MoffatField,CA 94035 ,USA
27SETIInstitute/NASA AmesResearchCenter,MS244-30,Moffa tField, CA 94035,USA– 3 –
ABSTRACT
Asteroseismology of stars in clusters has been a long-sough t goal because the as-
sumption of a common age, distance and initial chemical comp osition allows strong
tests of the theory of stellar evolution. We report results f rom the first 34 days of sci-
encedatafromthe KeplerMission fortheopenclusterNGC6819—oneoffourclus-
ters in the field of view. We obtain the first clear detections o f solar-like oscillations
in the cluster red giants and are able to measure the large fre quency separation, ∆ν,
andthefrequencyofmaximumoscillationpower, νmax. Wefindthattheasteroseismic
parameters allow us to test cluster-membership of the stars , and even with the limited
seismicdatainhand,wecan alreadyidentifyfourpossiblen on-membersdespitetheir
havinga betterthan 80% membershipprobabilityfrom radial velocitymeasurements.
We are also able to determine the oscillation amplitudes for stars that span about two
orders of magnitude in luminosity and find good agreement wit h the prediction that
oscillation amplitudesscale as the luminosityto the power of 0.7. These early results
demonstrate the unique potential of asteroseismology of th e stellar clusters observed
byKepler.
Subjectheadings: stars: fundamentalparameters—stars: oscillations—star s: interi-
ors—techniques: photometric—openclustersandassociati ons: individual(NGC6819)
1. Introduction
Openclustersprovideuniqueopportunitiesinastrophysic s. Starsinopenclustersarebelieved
to be formed from the same cloud of gas at roughly the same time . The fewer free parameters
available to model cluster stars make them interesting targ ets to analyze as a uniform ensemble,
especiallyforasteroseismicstudies.
Asteroseismology is an elegant tool based on the simple prin ciple that the frequency of a
standing acoustic wave inside a star depends on the sound spe ed, which in turn depends on
the physical properties of the interior. This technique app lied to the Sun (helioseismology) has
provided extremely detailed knowledge about the physics th at governs the solar interior, (e.g.,
Christensen-Dalsgaard2002). Allcoolstarsareexpectedt oexhibitsolar-likeoscillationsofstand-
ing acoustic waves – called p modes – that are stochastically driven by surface convection. Using
asteroseismology to probe the interiors of cool stars in clu sters, therefore, holds promise of re-
warding scientific return (Gough& Novotny 1993; Brown& Gill iland 1994). This potential has
resulted in several attempts to detect solar-like oscillat ions in clusters using time-series photome-
try. These attempts were often aimed at red giants, since the iroscillation amplitudesare expected– 4 –
tobelargerthanthoseofmain-sequenceorsubgiantstarsdu etomorevigoroussurfaceconvection.
Despite these attempts, only marginal detections have been attained so far, limited either by the
lengthofthetimeseriesusuallyachievablethroughobserv ationswiththe HubbleSpaceTelescope
(Edmonds& Gilliland 1996; Stello&Gilliland 2009) or by the difficulty in attaining high preci-
sion from ground-based campaigns (e.g., Gillilandetal. 19 93; Stelloet al. 2007; Frandsen et al.
2007).
InthisLetterwereportcleardetectionsofsolar-likeosci llationsinred-giantstarsintheopen
cluster NGC 6819 using photometry from NASA’s Kepler Mission (Borucki et al. 2009). This
cluster,oneoffourinthe Keplerfield, isabout2.5Gyrold. Itisatadistanceof2.3kpc, andha sa
metallicityof[Fe/H] ∼ −0.05(see Holeet al. 2009, and references herein).
2. Observations anddata reduction
The data were obtained between 2009 May 12 and June 14, i.e., t he first 34 days of con-
tinuous science observations by Kepler(Q1 phase). The spacecraft’s long-cadence mode ( ∆t≃
30minutes) used in this investigation provided a total of 1639 data points in the time series of
each observed star. For this Letter we selected 47 stars in th e field of the open cluster NGC 6819
with membership probability PRV>80% from radial velocity measurements (Holeet al. 2009).
Figure1showsthecolor-magnitudediagram(CMD)oftheclus terwiththeselectedstarsindicated
by green symbols. The eleven annotated stars form a represen tative subset, which we will use to
illustrate our analyses in Sections 3 and 4. We selected the s tars in this subset to cover the same
brightnessrangeasourfullsample,whilegivinghighweigh ttostarsthatappeartobephotometric
non-members (i.e., stars located far from the isochrone in t he CMD). Data for each target were
checked carefully to ensure that the time-series photometr y was not contaminated significantly
by other stars in the field, which could otherwise complicate the interpretation of the oscillation
signal.
Fourteen data points affected by the momentum dumping of the spacecraft were removed
from the time series of each star. In addition, we removed poi nts that showed a point-to-point
deviation greater than 4σ, whereσis the local rms of the point-to-point scatter within a 24 hou r
window. This process removed on average one data-point per t ime series. Finally, we removed a
linear trend from each time series and then calculated the di screte Fourier transform. The Fourier
spectraathighfrequencyhavemeanlevelsbelow5partsperm illion(ppm)inamplitude,allowing
usto search forlow-amplitudesolar-likeoscillations.– 5 –
3. Extractionofasteroseismicparameters
Figure 2 shows the Fourier spectra (in power) of 9 stars from o ur subset. These range from
thelowerred-giant branch to thetip ofthe branch (see Figur e1). The stars are sorted by apparent
magnitude, which for a cluster is indicative of luminosity, with brightest at the top. Note that the
redgiantsinNGC6819aresignificantlyfainter( 12/lessorsimilarV/lessorsimilar14)thanthesampleof Keplerfieldred
giants (8/lessorsimilarV/lessorsimilar12) studied by Beddinget al. (2010). Nevertheless, it is clear from Figure 2 that
we can detect oscillations for stars that span about two orde rs of magnitude in luminosity along
theclustersequence.
Weusedfourdifferentpipelines(Hekkeret al.2009a;Huber et al.2009a;Mathuret al.2009;
Mosser& Appourchaux 2009) to extract the average frequency separation between modes of the
same degree (the so-called large frequency separation, ∆ν). We have also obtained the frequency
of maximum oscillation power, νmax, and the oscillation amplitude. The measured values of ∆ν
are indicated by vertical dotted lines in Figure 2 centered o n the highest oscillation peaks near
νmax. While the stars in Figure 2, particularly in the lower panel s, show the regular series of
peaks expected for solar-likeoscillations,the limitedle ngth of the time-series datadoes not allow
such structureto be clearly resolved for the mostluminouss tars in our sample— thosewith νmax
/lessorsimilar20µHz. We do, however, see humps of excess power in the Fourier sp ectra (see Figure 2 star
no. 2 and 8) with νmaxand amplitude in mutual agreement with oscillations. With l onger time
series weexpectmorefirm resultsforthesehigh-luminosity giants.
4. Cluster membership from asteroseismology
It isimmediatelyclear fromFigure2thatnotallstars follo wtheexpected trendofincreasing
νmaxwith decreasing apparent magnitude, suggesting that some o f the stars might be intrinsically
brighterorfainterthanexpected. Sinceoscillationsinas taronlydependonthephysicalproperties
of the star, we can use asteroseismology to judge whether or n ot a star is likely to be a cluster
member independentlyof its distanceand of interstellarab sorption and reddening. For cool stars,
νmaxscaleswiththeacousticcut-offfrequency,anditiswelles tablishedthatwecanestimate νmax
by scalingfromthesolarvalue(Brownet al. 1991; Kjeldsen& Bedding 1995):
νmax
νmax,⊙=M/M⊙(Teff/Teff,⊙)3.5
L/L⊙, (1)
whereνmax,⊙= 3100µHz. The accuracy of such estimates is good to within 5% (Stell oet al.
2009)assumingwehavegoodestimatesofthestellarparamet ersM,L, andTeff.
In thefollowingweassumetheidealisticscenario whereall clustermembersfollowstandard
stellar evolutiondescribed by the isochrone. Stellar mass along the red giant branch of thecluster– 6 –
isochrone varies by less than 1%. The variation is less than 5 % even if we also consider the
asymptoticgiant branch. For simplicity,we therefore adop t a mass of 1.55M⊙for all stars, which
is representativefortheisochronefrom Marigoet al. (2008 )(Figure 1) and a similarisochroneby
VandenBerg etal. (2006). Neglectingbinarity (see Table 1) , we derivethe luminosityof each star
in our subset from its V-band apparent magnitude, adopting reddening and distance modulus of
E(B−V) = 0.1and(M−m)V= 12.3,respectively(obtainedfromsimpleisochronefitting,see
Holeetal.2009). WeusedthecalibrationofFlower(1996)to convertthestellar (B−V)0colorto
Teff. BolometriccorrectionswerealsotakenfromFlower(1996) . Thederivedquantitieswerethen
used toestimate νmaxfor each star(Eq.1), and compared withtheobservedvalue(s eeFigure3).
Figure 3 shows four obvious outliers (no. 1, 3, 8 and 11), thre e of which are also outliers in
theCMD (no. 1, 3, and11). Fortherest ofthestars weseegood a greement between theexpected
andobservedvalue,indicatingthattheuncertaintyonthe νmaxestimatesarerelativelysmall. Since
thevariationsinmassandeffectivetemperatureamongthec lustergiantstarsaresmall,deviations
fromthedottedlinemustbecausedbyanincorrectestimateo ftheluminosity. Thisimpliesthatthe
luminositiesofstarsfallingsignificantlyaboveorbelowt helinehavebeenover-orunderestimated,
respectively. The simplest interpretation is that these ou tliers are fore- or background stars, and
hence not members of the cluster. To explain the differences between the observed and expected
value ofνmaxwould require the deviant stars to have Verrors of more than 1 magnitude, and in
some cases B−Verrors of about 0.2 magnitude if they were cluster members. B inarity may
explain deviations above the dotted line, but only by up to a f actor of two in L(and hence, in the
ratio of the observed to expected νmax). The deviation of only one star (no.1) could potentially
be explained this way. However, that would be in disagreemen t with its single-star classification
from multi-epoch radial velocity measurements, assuming i t is not a binary viewed pole-on (see
Table 1). Hence, under the assumptionof a standard stellar e volution, the most likely explanation
forallfouroutliersinFigure3isthereforethatthesestar sarenotclustermembers. Thisconclusion
is, however,in disagreementwith theirhighmembershippro babilityfrom measurementsofradial
velocity (Holeet al. 2009) and proper motion (Sanders 1972) (see Table 1). Another interesting
possibility is that the anomalous pulsation properties mig ht be explained by more exotic stellar
evolutionscenariosthan isgenerally anticipatedforopen -clusterstars.
5. Asteroseismic“color-magnitude diagrams”
ItisclearfromFigure2thattheamplitudesoftheoscillati onsincreasewithluminosityforthe
seismicallydeterminedclustermembers. Basedoncalculat ionsbyChristensen-Dalsgaard& Frandsen
(1983), Kjeldsen& Bedding (1995) have suggested that the ph otometric oscillation amplitude of
p modes scale as (L/M)sTeff−2, withs= 1(the velocity amplitudes, meanwhile, would scale as– 7 –
(L/M)s). This was revised by Samadi etal. (2007) to s= 0.7based on models of main sequence
stars. Takingadvantageofthefewerfreeparameterswithin thisensembleofstars,ourobservations
allow us to make some progress towards extrapolating this sc aling to red giants and determining
thevalueof s.
In Figure4 weintroduceanewtypeofdiagramthatissimilart oaCMD, butwithmagnitude
replaced by an asteroseismicparameter – in thiscase, theme asured oscillationamplitude. Ampli-
tudeswereestimatedforallstarsinoursample(exceptfort hefouroutliers)usingmethodssimilar
tothatofKjeldsenet al.(2008)(seealsoMichelet al.2008) ,whichassumethattherelativepower
betweenradialandnon-radialmodesisthesameasintheSun. Thisdiagramconfirmstherelation-
ship between amplitude and luminosity. Despite a large scat ter, which is not surprising from this
relatively short timeseries, we see that s= 0.7provides a much better match than s= 1.0. Once
verifiedwithmoredata,thisrelationwillallowtheuseofth emeasuredamplitudeasanadditional
asteroseismic diagnostic for testing cluster membership a nd for isochrone fitting in general. We
notethat theother clusters observed by Keplerhave different metallicitiesthan NGC 6819, which
willallowfutureinvestigationon themetallicitydepende nce oftheoscillationamplitudes.
We expect to obtain less scatter in the asteroseismic measur ements when longer time series
become available. That will enable us to expand classical is ochrone fitting techniques to include
diagramslikethis,whereamplitudecouldalsobereplacedb yνmaxor∆ν. Inparticular,weshould
beabletodeterminetheabsoluteradiiaidedby ∆νoftheredgiantbranchstars,whichwouldbean
importantcalibratorfor theoretical isochrones. Additio nally,thedistributionsoftheasteroseismic
parameters – such as νmax– can potentially be used to test stellar population synthes is models
(Hekkeret al.2009b;Miglioet al.2009b). Applyingthisapp roachtoclusterscouldleadtofurther
progress in understanding of physical processes such as mas s loss during the red-giant phase (see
e.g.,Miglioet al.2009a). Notethatafewclearoutliersare indicativeofnon-membershiporexotic
stellarevolution,asaresultoffactorssuchasstellarcol lisionsorheavymassloss,whileageneral
deviationfromthetheoreticalpredictionsbyalargegroup ofstarswouldsuggestthatthestandard
theorymay need revision.
Finally, we note that NGC 6819 and another Keplercluster, NGC 6791, contain detached
eclipsingbinaries(Talamantes& Sandquist2009;Street et al.2005;deMarchi et al.2007;Mochejskaetal.
2005). For these stars masses and radii can be determined ind ependently (Grundahl et al. 2008),
whichwillfurtherstrengthenresultsofasteroseismicana lyses.– 8 –
6. Discussion& Conclusions
PhotometricdataofredgiantsinNGC6819obtainedbyNASA’s KeplerMission haveenabled
ustomakethefirst cleardetectionofsolar-likeoscillatio nsin clusterstars. Thegeneral properties
of the oscillations ( ∆ν,νmax, and amplitudes) agree well with results of field red giants m ade by
Kepler(Bedding etal.2010)andCoRoT(deRidderet al.2009;Hekker et al.2009b). Wefindthat
the oscillation amplitudes of the observed stars scale as (L/M)0.7Teff−2, suggesting that previous
attemptstodetect oscillationsinclustersfrom groundwer eat thelimitofdetection.
We find that the oscillation properties provide additional t ests for cluster membership, al-
lowing us to identify four stars that are either non-members or exotic stars. All four stars have
membership probability higher than 80% from radial-veloci ty measurements, but three of them
appear to be photometric non-members. We further point out t hat deviations from the theoretical
predictionsoftheasteroseismicparametersamongalarges ampleofclusterstarshavethepotential
ofbeingusedasadditionalconstraintsintheisochronefitt ingprocess,whichcanleadtoimproved
stellarmodels.
Our results, based on limited data of about one month, highli ght the unique potential of as-
teroseismologyon the brighteststars in thestellarcluste rs observed by Kepler. With longerseries
sampled at the spacecraft’s short cadence ( ≃1 minute), we expect to detect oscillations in the
subgiantsand turn-offstars, as wellas inthebluestraggle rsinthiscluster.
FundingforthisDiscoverymissionisprovidedbyNASA’sSci enceMissionDirectorate. The
authorswouldliketothanktheentire Keplerteamwithoutwhomthisinvestigationwouldnothave
been possible. The authors also thank all funding councils a nd agencies that have supported the
activitiesofWorkingGroup 2ofthe KeplerAsteroseismicScience Consortium(KASC).
Facilities: Kepler.
REFERENCES
Bedding,T. R., et al. 2010,ApJL,inpress
Borucki, W.,et al. 2009,inIAU Symposium,Vol.253, IAUSymp osium,289
Brown, T.M.,& Gilliland,R. L. 1994,ARA&A,32, 37
Brown, T.M.,Gilliland,R. L., Noyes,R. W.,& Ramsey,L. W.19 91,ApJ, 368,599
Christensen-Dalsgaard,J.2002,ReviewsofModern Physics ,74, 1073– 9 –
Christensen-Dalsgaard,J.,& Frandsen, S. 1983,Sol. Phys. ,82,469
deMarchi,F., etal. 2007,A&A,471, 515
deRidder, J.,et al. 2009,Nature, 459,398
Edmonds,P. D., &Gilliland,R. L.1996, ApJ,464,L157
Flower, P. J.1996,ApJ, 469,355
Frandsen, S., et al. 2007,A&A,475,991
Gilliland,R. L., et al. 1993,AJ,106,2441
Gough, D. O., & Novotny, E. 1993, in ASP Conf. Ser. 42: GONG 199 2. Seismic Investigationof
theSunand Stars, ed. T.M. Brown,355
Grundahl,F., Clausen, J. V.,Hardis, S., &Frandsen, S. 2008 ,A&A,492,171
Hekker, S., et al. 2009a,MNRAS, in press(astro-ph/0911.26 12)
—.2009b,A&A, 506,465
Hole, K. T., Geller, A. M., Mathieu, R. D., Platais, I., Meibo m, S., & Latham, D. W. 2009, AJ,
138,159
Huber, D., Stello, D., Bedding, T. R., Chaplin, W. J., Arento ft, T., Quirion, P., & Kjeldsen, H.
2009a,Commun.Asteroseismol.,160,74
Kjeldsen,H., &Bedding,T. R. 1995,A&A,293, 87
Kjeldsen,H., etal. 2008,ApJ, 682,1370
Latham,D. W.,Brown, T.M.,Monet,D. G., Everett,M.,Esquer do,G. A.,& Hergenrother, C. W.
2005,inBulletinoftheAmerican AstronomicalSociety,Vol . 37,1340
Marigo, P., Girardi, L., Bressan, A., Groenewegen, M. A. T., Silva, L., & Granato, G. L. 2008,
A&A,482,883
Mathur,S., et al. 2009,A&A,inpress (arXiv:0912.3367)
Michel,E., etal. 2008,Science, 322,558
Miglio, A., Montalb´ an, J., Eggenberger, P., Hekker, S., & N oels, A. 2009a, in American Institute
ofPhysicsConference Series, Vol.1170,AmericanInstitut eofPhysicsConference Series,
ed. J.A. Guzik& P. A. Bradley,132– 10 –
Miglio,A., et al.2009b,A&A,503, L21
Mochejska,B. J., et al.2005,AJ, 129,2856
Mosser,B., & Appourchaux,T.2009,A&A,508, 877
Samadi, R., Georgobiani, D., Trampedach, R., Goupil, M. J., Stein, R. F., & Nordlund, ˚A. 2007,
A&A,463,297
Sanders, W. L. 1972,A&A,19,155
Stello,D., Chaplin,W. J.,Basu, S., Elsworth,Y., &Bedding , T.R. 2009, MNRAS, 400,80
Stello,D., &Gilliland,R. L.2009,ApJ, 700,949
Stello,D., et al. 2007,MNRAS, 377,584
Street, R. A.,et al. 2005,MNRAS, 358,795
Talamantes, A., & Sandquist, E. L. 2009, in Bulletin of the Am erican Astronomical Society,
Vol.41,320
VandenBerg, D. A., Bergbusch, P. A., &Dowler,P. D. 2006,ApJ S, 162,375
ThispreprintwaspreparedwiththeAAS L ATEXmacrosv5.2.– 11 –
Table1:Cross identificationsandmembership.
ID ID WOCS ID ID Mem.ship Mem.ship Mem.ship
Thiswork KICaHoleet al. Sanders Holeet al.bSanderscThiswork
1 5024272 003003 SM95% no
2 5024750 001004 141 SM93% 83% yes
3 5023889 004014 42 SM95% 90% no
4 5023732 005014 27 SM94% 90% yes
5 5112950 003005 148 SM95% 92% yes
6 5112387 003007 73 SM95% 88% yes
7 5024512 003001 116 SM93% 90% yes
8 4936335 007021 9 SM95% 68% no
9 5024405 004001 100 SM93% 91% yes
10 5112072 009010 39 SM95% 91% yes
11 4937257 009015 144 SM88% 80% no
aIDfromthe KeplerInputCatalogue (Lathamet al. 2005).
bClassification (SM:singlemember)andmembershipprobabil ityfromradialvelocity(Holeetal. 2009).
cMembershipprobabilityfrompropermotion(Sanders1972).– 12 –
Fig. 1.— Color-magnitude diagram of NGC 6819. Plotted stars have membership probability
PRV>80% as determined by Holeet al. (2009). Photometric indices ar e from the same source.
Theisochroneis from Marigoet al. (2008)(Age=2.4 Gyr, Z=0. 019,modified for theadopted red-
dening of 0.1mag). Color-coded stars have been analyzed, an d the annotated numbers refer to the
legend in panels of Figure 2 and star numbers in Figure 3 (see a lso Table 1). Insets show light
curves in parts per thousand of two red giants oscillating on different timescales. The variations
ofthelightcurves inPanelA and Baredominatedby thestella roscillationswithperiodsofafew
days andofaboutsix hours,respectively.– 13 –
Fig. 2.— Fourierspectraofa representativeset ofred giant salongtheclustersequence sortedby
apparent magnitude. Annotated numbers in each panel refer t o the star identification (see Fig. 1
and Table 1). ‘AM’ indicates that the star is an asteroseismi c member. Red solid curves show the
smoothed spectrum for stars with νmax<20µHz. To guide the eye, we have plotted dotted lines
toindicatethemeasuredaveragelargefrequencyseparatio n. Thecentraldottedlineiscenteredon
thehighestoscillationpeaksnear νmax. Notethatsince ∆νisgenerallyfrequencydependent,only
thecentraldottedlineisexpectedtolineupwithapeakinth eoscillationspectrum. Theredarrows
indicate the position of the expected νmax(see Eq. 1) for stars where the observed value does not
agree withtheexpectationsforthiscluster(seeSection 4) .– 14 –
Fig. 3.— Ratioofobservedandexpected νmax. 1-σerrorbarsindicatetheuncertaintyon νmax(obs).
Stars clearly above or below the dotted line are either not cl uster members or members whose
evolutionhavenot followedthestandardscenario.– 15 –
Fig. 4.— Amplitude color diagram of red giant stars in NGC 681 9 with the Marigoet al. (2008)
isochrone overlaid with three values of sin the amplitude scaling relation: (L/M)sTeff−2. The
solarvalueusedin thisscalingis 4.7ppm(Kjeldsen &Bedding 1995).