Panchromatic fits to the Globular Cluster NGC 6366

0.98 MB·English

by  Fabíola Campos

Mem.S.A.It.Vol.75,282
(cid:13)c SAIt 2008 Memoriedella
Panchromatic fits to the Globular Cluster NGC
6366
Fab´ıola Campos1,S.O. Kepler1 andC. Bonatto1
3
1
0 1 Instituto de F´ısica – Universidade Federal do Rio Grande do Sul, Avenida Bento
2 Gonc¸alves9500,91509-900PortoAlegre,RioGrandedoSul,Brasil
e-mail:[email protected]
n
a
J
8 Abstract. Wepresentpanchromaticisochronefitstothecolormagnitudedataoftheglob-
1 ular cluster NGC 6366, based on HST ACS/WFC and SOAR photometric data. Before
performing the isochrone fits, we corrected the photometric data for differential redden-
] ing and calculated the mean ridge line of the color magnitude diagrams. We compared
R the isochrones of Dartmouth StellarEvolution Database and PAdova and TRiesteStellar
S EvolutionCode(withmicroscopicdiffusionstartingonthemainsequence).Basedonprevi-
. ousdeterminationsofthemetallicityofthisclusterwetestitfrom[Fe/H]=-1.00to[Fe/H]=
h -0.50,andtheagefrom9to13Gyrs.Theuncertaintiesdonotdecreasewhenwefitsimulta-
p neouscolors.WealsofindthattheDartmouthStellarEvolutionDatabaseisochroneshavea
- betterfitinthesubgiantbranchandlowmainsequencethanthePAdovaandTRiesteStellar
o
EvolutionCode.Consideringthemostrecentspectroscopicdeterminationofthemetallicity
r
t ([Fe/H]=-0.67),wefindE(B-V)=0.67±0.02,(m-M) =14.94±0.05and11±2GyrforNGC
s V
6366.
a
[
Key words. Globular Clusters: General – Globular Clusters: Individual – Stars: Color

2 MagnitudeDiagram–Galaxy:FundamentalParameters
v
0
4
1. Introduction eters that can vary to generate the models to
1
4 becomparedwiththedata,ageandmetallicity,
Galactic globular clusters are considered the
. andtwootherfittingparameters,extinctionand
1 ideal laboratories for the study of stellar evo-
distance.Onemethodcommonlyusedisthefit
0 lution, mainly because their color magnitude
”byeye”,butitisnecessarytobeawareofthe
3 diagram(CMDs)haveveryspecificcharacter-
1 uncertaintiesintheconstructionofisochrones,
istics.Thestars,inmostglobularclusters,fol-
: sincetheyincreasetheuncertaintiesinthefit.
v low a single isochrone, suggesting that they
Amongtheproblemsoftheevolutionarymod-
i formed rougly at the same time and with the
X els are the lack of a precise description of
samemetallicity.
convection.Redgiantstarshavedeepconvec-
r To obtain the astrophysical parameters of the
a tiveenvelopecausingalargeuncertaintyinthe
globular cluster is necessary to fit models to
structure. Another problem is that stars lose
the stars in the CMD. But this fitting is not
mass(higherforthemostmassivestars)inthe
simple,becausetherearetwophysicalparam-
form of a stellar wind and this loss increases
several orders of magnitude for stars that al-
Sendoffprintrequeststo:Fab´ıolaCampos
Fab´ıolaCampos:PanchromaticfitstotheglobularclusterNGC6366 283
ready left the main sequence. Predicting the gc.ca/community/STETSON/standards/).
rateofmasslosstheoreticallyisverydifficult, The photometric data of HST ACS/WFC was
and what the evolutionary models use are the obtained from http://www.astro.ufl.
values of mass loss consistent with observa- edu/˜ata/public_hstgc/. This data was
tions of stars that are at a similar stage. This obtained as partof the HST treasury program
value dependsheavily on metallicity, and this “An ACS Survey of the Galactic Globular
dependenceis difficult of measurement,caus- Clusters”[GO10775 P.I. Ata Sarajedini,
ing uncertainties. It is still necessary to ac- (Sarajedini et al. 2007)]. The images were
counttheuncertaintiesassociatedwithopacity centered at the cluster core and consists
tables, especially when dealing with molecu- of 10x(140s) and 1x(10s) exposures in the
lar opacities.This effectis importantnotonly F606WandF814Wfilters.
in the coolergiant stars, but also in the lower
mainsequence.
3. DataAnalysis
According to Bolte & Hogan (1995), if the
brightnessofthemainsequenceturnoffisused
Toperformtheanalysis,thefirststepinvolves
to determine the age of globular clusters, the correctingthephotometricdatafordifferential
only source of uncertainty significantly large
reddeningbydividingthefieldofviewacross
is the determinationof distance andan uncer-
theclusterinaregularcellgrid,thenextractthe
tainty of 25% in distance generates an uncer-
Hessdiagramfromeachcell,shiftingit,along
taintyof22%inage.
thereddeningvector,untilitmatchesthemean
Attempting to decrease the uncertainties of
diagram [Bonatto, Campos & Kepler (2013),
isochronefittingsto globularclusters, we per-
MNRAS,submitted].
formed panchromatic isochrone fits of the Totakeintoaccounttheeffectsofbinarityand
globularclusterNGC 6366,based onarchival
alsothescatteringphotometry,beforeperform-
HSTACS/WFCandourown4.1mSOARpho-
inganyfittingtotheCMDs,wecalculatedthe
tometricdata.We comparedtheisochronesof
meanridgelineofeachCMD.Thefittingofthe
DartmouthStellarEvolutionDatabase[DSED
DSEDandPARSECmodelstothemeanridge
-Dotteretal.(2008)]andPAdovaandTRieste
line of the three colors (VxB-V, VxU-V and
StellarEvolutionCode[PARSEC-Bressanet
F606WxF606W-F814W) was performed con-
al.(2012)],modelswithmicroscopicdiffusion
sideringalltherangeofmetallicitypreviously
starting on the main sequence (Jofre´ & Weiss determinedforthiscluster(-1.0<[Fe/H]<-0.5,
2011).
Ferroetal.(2008)andreferencestherein)and
ages ranging from 9 Gyr to 13 Gyr. In Fig. 1
we show the fitting of DSED models to the
2. Observations
mean ridge line at U-V color. It is not diffi-
The optical data of NGC 6366 discussed cult to notice, by looking at the low main se-
in this work was observed with the SOAR quence and the sub giant branch that, as the
telescope. The images were centered on the metallicity decreases, the models have a bet-
cluster core and the exposure times for each terfittothedata,until[Fe/H]=-0.67,whenthe
filter were divided as follows: 6x(1800s) for best fit is found, this value is consistent with
U; 5x(30s), 2x(300s) and 1x(1800s) for B; thespectroscopicdeterminationbyDaCosta&
and 5x(30s), 2x(300s) and 2x(1800s) for V . Armandroff(1995).Asmetallicitycontinuesto
The photometry was obtained with psf fitting decrease,themodelsnolongerfitthedata.
using DAOPHOT (Stetson 1987). To obtain The fitting for the three colors with
the magnitudes in the standard photometric [Fe/H]=-0.67±0.07 (Da Costa & Armandroff
system, stars belonging to the cluster itself 1995) (Fig. 2) shows that while the best fit to
were used. the standard magnitudes for these U-V coloroccurs at age 9 Gyr, forB-V color
stars were obtained through the catalog thebestfitisat11GyrandforF606W-F814W
of standard stars of Peter Stetson (http: colorthebestfitisfoundwith13Gyr.Inother
//www3.cadc-ccda.hia-iha.nrc-cnrc. words, a single model does not fit the three
284 Fab´ıolaCampos:PanchromaticfitstotheglobularclusterNGC6366
14 14 14
16 16 16
18 18 18
20 20 20
22 22 22
1 2 3 4 1 2 3 4 1 2 3 4
14 14 14
16 16 16
18 18 18
20 20 20
22 22 22
1 2 3 4 1 2 3 4 1 2 3 4
Fig.1. Fitting of the DSED to the mean ridge line (red dash-dotted line) of NGC 6366, at U-V color,
consideringalltherangeofmetallicitydeterminationsforthisclusterandagesof9Gyr(blueline),11Gyr
(magentaline)and13Gyr(greenline).Thebestfitisfoundatmetallicity[Fe/H]=-0.67,valueconsistent
withthemostrecentspectroscopicdeterminationbyDaCosta&Armandroff(1995).
colorssimultaneously,andtheuncertaintiesdo tersofthecluster.Beforewefindtherelations
notdecreasewhensimultaneouscolorsarefit- between the colors, to obtain the parameters,
ted.Indicatingthattheevoluionarymodelsstill we determined the total to selective extinc-
have unsolved problems, such as photometric tionforstarsbelongingtothisclusterdirectly,
zeropointsandopacitytables. followingDucati,Ribeiro&Rembold(2003),
AnotherinterestingpointtonoticeinFig.2is and we find R =3.06±0.14. We also used
V
the factthatDSED modelshavea betterfit at Cardelli,Clayton&Mathis(1989)relationsto
the lower main sequence and at the sub giant estimate E(B-V)=1.02E(F606W-F814W) and
branchthanthePARSECmodels;thereforewe E(B-V)=0.57E(U-V). We find the following
usetheDSEDfittingstodeterminetheparame- parametersforNGC6366:
Fab´ıolaCampos:PanchromaticfitstotheglobularclusterNGC6366 285
E(B-V)=0.69±0.02(int)±0.04(ext)
(m-M) =15.02±0.07(int)±0.13(ext)
V 14 14
Age=11±2Gyr.
4. Conclusions 16 16
TheuncertaintiesofisochronefittingstoNGC
6366donotdecreasewhenwefitmultiplecol-
18 18
ors,duetothefactthatasinglemodeldoesnot
fitthreecolorssimultaneously.Itindicatesthat
the isochronemodelsmaystill have problems
20 20
that remain unsolved, possibly photometric
zeropointsandopacitytables.Otherimportant
point is that the DSED models have a better 22 22
fit to the data than PARSEC models, mainly
in the sub giant branch and the low main 1 2 3 4 1 2 3 4
sequence, the last one is possibly realted to
the equation of state adopted by Dotter et al.
(2008)forstarswithlowmass. 14 14
With the total to selective relation
R =3.06±0.14, determined for stars be-
V
longing to this cluster, we estimated 16 16
E(B-V)=1.02E(F606W-F814W) and E(B-
V)=0.57E(U-V). With that we determined
the parameters of NGC 6366 as being: E(B- 18 18
V)= 0.69±0.02(int)±0.04(ext); (m-M) =
V
15.02±0.07(int)±0.13(ext) and Age= 11±2
Gyr. 20 20
Acknowledgements. Finacial support for this
researchcomesfromNationalCouncilforScientific
22 22
and Technological Development (CNPq) and
PRONEX-FAPERGS/CNPq.
1 1.2 1.4 1.6 1.8 2 1 1.2 1.4 1.6 1.8 2
References
Bolte, M. & Hogan, C. J. 1995, Nature, 376,
399
Bressan,A.etal.2012,MNRAS,427,127 15 15
Cardelli, J. A. Clayton, G. C. Mathis, J. S.
1989, IAUS,135,5
DaCosta,G.S.,&Armandroff,T.E.1995,AJ,
109,2533
Dotter,A.etal.2008,ApJS,178,89
20 20
Ducati, J. R. Ribeiro, D. , Rembold, S. B.
1998, ApJ,588,344
Ferro,A.A.etal.2008,RMxAA,44,365
Jofre´, M. & Weiss, C. J. 2011, A&A, 533,
id.A59
Sarajedini,A.etal.2007, AJ,133,1658 25 25
1 1.2 1.4 1.6 1.8 2 1 1.2 1.4 1.6 1.8 2
Stetson,P.B.1987,PASP,99,191
Fig.2. Fitting of PARSEC and DSED models to
the mean ridge line (red dash-dotted line) of the
SOARandACS/WFCdata,with[Fe/H]=0.67±0.07
(DaCosta&Armandroff 1995) and ages of 9Gyr
(blueline),11Gyr(magentaline)and13Gyr(green
line).Theuncertaintiesdonotdecreasewhenwefit
simultaneous colors, because a single model does
not fit the three colors simultaneously. The DSED
models haveabetterfitintheHertzprung gapand
thelowermainsequencethanthePARSECmodels.